Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux

Pull s390 updates from Martin Schwidefsky:

- Add the CPU id for the new z13s machine

- Add a s390 specific XOR template for RAID-5 checksumming based on the
XC instruction. Remove all other alternatives, XC is always faster

- The merge of our four different stack tracers into a single one

- Tidy up the code related to page tables, several large inline
functions are now out-of-line. Bloat-o-meter reports ~11K text size
reduction

- A binary interface for the priviledged CLP instruction to retrieve
the hardware view of the installed PCI functions

- Improvements for the dasd format code

- Bug fixes and cleanups

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux: (31 commits)
s390/pci: enforce fmb page boundary rule
s390: fix floating pointer register corruption (again)
s390/cpumf: add missing lpp magic initialization
s390: Fix misspellings in comments
s390/mm: split arch/s390/mm/pgtable.c
s390/mm: uninline pmdp_xxx functions from pgtable.h
s390/mm: uninline ptep_xxx functions from pgtable.h
s390/pci: add ioctl interface for CLP
s390: Use pr_warn instead of pr_warning
s390/dasd: remove casts to dasd_*_private
s390/dasd: Refactor dasd format functions
s390/dasd: Simplify code in format logic
s390/dasd: Improve dasd format code
s390/percpu: remove this_cpu_cmpxchg_double_4
s390/cpumf: Improve guest detection heuristics
s390/fault: merge report_user_fault implementations
s390/dis: use correct escape sequence for '%' character
s390/kvm: simplify set_guest_storage_key
s390/oprofile: add z13/z13s model numbers
s390: add z13s model number to z13 elf platform
...

arch/s390/Kconfig

···
 	  older machines.
 
 config MARCH_Z13
-	bool "IBM z13"
+	bool "IBM z13s and z13"
 	select HAVE_MARCH_Z13_FEATURES
 	help
-	  Select this to enable optimizations for IBM z13 (2964 series).
-	  The kernel will be slightly faster but will not work on older
-	  machines.
+	  Select this to enable optimizations for IBM z13s and z13 (2965 and
+	  2964 series). The kernel will be slightly faster but will not work on
+	  older machines.
 
 endchoice
 

arch/s390/include/asm/clp.h

···
 /* CLP common request & response block size */
 #define CLP_BLK_SIZE			PAGE_SIZE
 
+#define CLP_LPS_BASE	0
+#define CLP_LPS_PCI	2
+
 struct clp_req_hdr {
 	u16 len;
 	u16 cmd;
+	u32 fmt		: 4;
+	u32 reserved1	: 28;
+	u64 reserved2;
 } __packed;
 
 struct clp_rsp_hdr {
 	u16 len;
 	u16 rsp;
+	u32 fmt		: 4;
+	u32 reserved1	: 28;
+	u64 reserved2;
 } __packed;
 
 /* CLP Response Codes */
···
 #define CLP_RC_RESNOT0			0x0070	/* Reserved field not zero */
 #define CLP_RC_NODATA			0x0080	/* No data available */
 #define CLP_RC_FC_UNKNOWN		0x0100	/* Function code not recognized */
+
+/* Store logical-processor characteristics request */
+struct clp_req_slpc {
+	struct clp_req_hdr hdr;
+} __packed;
+
+struct clp_rsp_slpc {
+	struct clp_rsp_hdr hdr;
+	u32 reserved2[4];
+	u32 lpif[8];
+	u32 reserved3[8];
+	u32 lpic[8];
+} __packed;
+
+struct clp_req_rsp_slpc {
+	struct clp_req_slpc request;
+	struct clp_rsp_slpc response;
+} __packed;
 
 #endif

arch/s390/include/asm/gmap.h

···
+/*
+ *  KVM guest address space mapping code
+ *
+ *    Copyright IBM Corp. 2007, 2016
+ *    Author(s): Martin Schwidefsky <[e-mail]>
+ */
+
+#ifndef _ASM_S390_GMAP_H
+#define _ASM_S390_GMAP_H
+
+/**
+ * struct gmap_struct - guest address space
+ * @crst_list: list of all crst tables used in the guest address space
+ * @mm: pointer to the parent mm_struct
+ * @guest_to_host: radix tree with guest to host address translation
+ * @host_to_guest: radix tree with pointer to segment table entries
+ * @guest_table_lock: spinlock to protect all entries in the guest page table
+ * @table: pointer to the page directory
+ * @asce: address space control element for gmap page table
+ * @pfault_enabled: defines if pfaults are applicable for the guest
+ */
+struct gmap {
+	struct list_head list;
+	struct list_head crst_list;
+	struct mm_struct *mm;
+	struct radix_tree_root guest_to_host;
+	struct radix_tree_root host_to_guest;
+	spinlock_t guest_table_lock;
+	unsigned long *table;
+	unsigned long asce;
+	unsigned long asce_end;
+	void *private;
+	bool pfault_enabled;
+};
+
+/**
+ * struct gmap_notifier - notify function block for page invalidation
+ * @notifier_call: address of callback function
+ */
+struct gmap_notifier {
+	struct list_head list;
+	void (*notifier_call)(struct gmap *gmap, unsigned long gaddr);
+};
+
+struct gmap *gmap_alloc(struct mm_struct *mm, unsigned long limit);
+void gmap_free(struct gmap *gmap);
+void gmap_enable(struct gmap *gmap);
+void gmap_disable(struct gmap *gmap);
+int gmap_map_segment(struct gmap *gmap, unsigned long from,
+		     unsigned long to, unsigned long len);
+int gmap_unmap_segment(struct gmap *gmap, unsigned long to, unsigned long len);
+unsigned long __gmap_translate(struct gmap *, unsigned long gaddr);
+unsigned long gmap_translate(struct gmap *, unsigned long gaddr);
+int __gmap_link(struct gmap *gmap, unsigned long gaddr, unsigned long vmaddr);
+int gmap_fault(struct gmap *, unsigned long gaddr, unsigned int fault_flags);
+void gmap_discard(struct gmap *, unsigned long from, unsigned long to);
+void __gmap_zap(struct gmap *, unsigned long gaddr);
+void gmap_unlink(struct mm_struct *, unsigned long *table, unsigned long vmaddr);
+
+void gmap_register_ipte_notifier(struct gmap_notifier *);
+void gmap_unregister_ipte_notifier(struct gmap_notifier *);
+int gmap_ipte_notify(struct gmap *, unsigned long start, unsigned long len);
+
+#endif /* _ASM_S390_GMAP_H */

arch/s390/include/asm/pci.h

···
 	u64 rpcit_ops;
 	u64 dma_rbytes;
 	u64 dma_wbytes;
-} __packed __aligned(16);
+} __packed __aligned(64);
 
 enum zpci_state {
 	ZPCI_FN_STATE_RESERVED,
···
 
 /* Private data per function */
 struct zpci_dev {
-	struct pci_dev	*pdev;
 	struct pci_bus	*bus;
 	struct list_head entry;		/* list of all zpci_devices, needed for hotplug, etc. */
 
···
 /* Debug */
 int zpci_debug_init(void);
 void zpci_debug_exit(void);
-void zpci_debug_init_device(struct zpci_dev *);
+void zpci_debug_init_device(struct zpci_dev *, const char *);
 void zpci_debug_exit_device(struct zpci_dev *);
 void zpci_debug_info(struct zpci_dev *, struct seq_file *);
 

arch/s390/include/asm/pci_clp.h

···
 /* List PCI functions request */
 struct clp_req_list_pci {
 	struct clp_req_hdr hdr;
-	u32 fmt			:  4;	/* cmd request block format */
-	u32			: 28;
-	u64 reserved1;
 	u64 resume_token;
 	u64 reserved2;
 } __packed;
···
 /* List PCI functions response */
 struct clp_rsp_list_pci {
 	struct clp_rsp_hdr hdr;
-	u32 fmt			:  4;	/* cmd request block format */
-	u32			: 28;
-	u64 reserved1;
 	u64 resume_token;
 	u32 reserved2;
 	u16 max_fn;
···
 /* Query PCI function request */
 struct clp_req_query_pci {
 	struct clp_req_hdr hdr;
-	u32 fmt			:  4;	/* cmd request block format */
-	u32			: 28;
-	u64 reserved1;
 	u32 fh;				/* function handle */
 	u32 reserved2;
 	u64 reserved3;
···
 /* Query PCI function response */
 struct clp_rsp_query_pci {
 	struct clp_rsp_hdr hdr;
-	u32 fmt			:  4;	/* cmd request block format */
-	u32			: 28;
-	u64			: 64;
 	u16 vfn;			/* virtual fn number */
 	u16			:  7;
 	u16 util_str_avail	:  1;	/* utility string available? */
···
 /* Query PCI function group request */
 struct clp_req_query_pci_grp {
 	struct clp_req_hdr hdr;
-	u32 fmt			:  4;	/* cmd request block format */
-	u32			: 28;
-	u64 reserved1;
-	u32			: 24;
+	u32 reserved2		: 24;
 	u32 pfgid		:  8;	/* function group id */
-	u32 reserved2;
-	u64 reserved3;
+	u32 reserved3;
+	u64 reserved4;
 } __packed;
 
 /* Query PCI function group response */
 struct clp_rsp_query_pci_grp {
 	struct clp_rsp_hdr hdr;
-	u32 fmt			:  4;	/* cmd request block format */
-	u32			: 28;
-	u64 reserved1;
 	u16			:  4;
 	u16 noi			: 12;	/* number of interrupts */
 	u8 version;
···
 /* Set PCI function request */
 struct clp_req_set_pci {
 	struct clp_req_hdr hdr;
-	u32 fmt			:  4;	/* cmd request block format */
-	u32			: 28;
-	u64 reserved1;
 	u32 fh;				/* function handle */
 	u16 reserved2;
 	u8 oc;				/* operation controls */
···
 /* Set PCI function response */
 struct clp_rsp_set_pci {
 	struct clp_rsp_hdr hdr;
-	u32 fmt			:  4;	/* cmd request block format */
-	u32			: 28;
-	u64 reserved1;
 	u32 fh;				/* function handle */
 	u32 reserved3;
 	u64 reserved4;

arch/s390/include/asm/percpu.h

···
 	ret__;								\
 })
 
-#define this_cpu_cmpxchg_double_4 arch_this_cpu_cmpxchg_double
 #define this_cpu_cmpxchg_double_8 arch_this_cpu_cmpxchg_double
 
 #include <asm-generic/percpu.h>

arch/s390/include/asm/perf_event.h

···
 #define PMU_F_ERR_LSDA			0x0200
 #define PMU_F_ERR_MASK			(PMU_F_ERR_IBE|PMU_F_ERR_LSDA)
 
-/* Perf defintions for PMU event attributes in sysfs */
+/* Perf definitions for PMU event attributes in sysfs */
 extern __init const struct attribute_group **cpumf_cf_event_group(void);
 extern ssize_t cpumf_events_sysfs_show(struct device *dev,
 				       struct device_attribute *attr,

arch/s390/include/asm/pgalloc.h

···
 void page_table_free_rcu(struct mmu_gather *, unsigned long *, unsigned long);
 extern int page_table_allocate_pgste;
 
-int set_guest_storage_key(struct mm_struct *mm, unsigned long addr,
-			  unsigned long key, bool nq);
-unsigned long get_guest_storage_key(struct mm_struct *mm, unsigned long addr);
-
 static inline void clear_table(unsigned long *s, unsigned long val, size_t n)
 {
 	typedef struct { char _[n]; } addrtype;

arch/s390/include/asm/pgtable.h

···
 
 /*
  * Segment table entry encoding (R = read-only, I = invalid, y = young bit):
- *				dy..R...I...wr
+ *				dy..R...I...rw
  * prot-none, clean, old	00..1...1...00
  * prot-none, clean, young	01..1...1...00
  * prot-none, dirty, old	10..1...1...00
  * prot-none, dirty, young	11..1...1...00
- * read-only, clean, old	00..1...1...01
- * read-only, clean, young	01..1...0...01
- * read-only, dirty, old	10..1...1...01
- * read-only, dirty, young	11..1...0...01
+ * read-only, clean, old	00..1...1...10
+ * read-only, clean, young	01..1...0...10
+ * read-only, dirty, old	10..1...1...10
+ * read-only, dirty, young	11..1...0...10
  * read-write, clean, old	00..1...1...11
  * read-write, clean, young	01..1...0...11
  * read-write, dirty, old	10..0...1...11
···
 	return (pmd_val(pmd) & ~_SEGMENT_ENTRY_BITS) != 0;
 }
 
-#define  __HAVE_ARCH_PMDP_SET_ACCESS_FLAGS
-extern int pmdp_set_access_flags(struct vm_area_struct *vma,
-				 unsigned long address, pmd_t *pmdp,
-				 pmd_t entry, int dirty);
-
-#define __HAVE_ARCH_PMDP_CLEAR_YOUNG_FLUSH
-extern int pmdp_clear_flush_young(struct vm_area_struct *vma,
-				  unsigned long address, pmd_t *pmdp);
-
 #define __HAVE_ARCH_PMD_WRITE
 static inline int pmd_write(pmd_t pmd)
 {
···
 {
 	pmd_val(pmd) &= ~_SEGMENT_ENTRY_SOFT_DIRTY;
 	return pmd;
-}
-
-static inline pgste_t pgste_get_lock(pte_t *ptep)
-{
-	unsigned long new = 0;
-#ifdef CONFIG_PGSTE
-	unsigned long old;
-
-	preempt_disable();
-	asm(
-		"	lg	%0,%2\n"
-		"0:	lgr	%1,%0\n"
-		"	nihh	%0,0xff7f\n"	/* clear PCL bit in old */
-		"	oihh	%1,0x0080\n"	/* set PCL bit in new */
-		"	csg	%0,%1,%2\n"
-		"	jl	0b\n"
-		: "=&d" (old), "=&d" (new), "=Q" (ptep[PTRS_PER_PTE])
-		: "Q" (ptep[PTRS_PER_PTE]) : "cc", "memory");
-#endif
-	return __pgste(new);
-}
-
-static inline void pgste_set_unlock(pte_t *ptep, pgste_t pgste)
-{
-#ifdef CONFIG_PGSTE
-	asm(
-		"	nihh	%1,0xff7f\n"	/* clear PCL bit */
-		"	stg	%1,%0\n"
-		: "=Q" (ptep[PTRS_PER_PTE])
-		: "d" (pgste_val(pgste)), "Q" (ptep[PTRS_PER_PTE])
-		: "cc", "memory");
-	preempt_enable();
-#endif
-}
-
-static inline pgste_t pgste_get(pte_t *ptep)
-{
-	unsigned long pgste = 0;
-#ifdef CONFIG_PGSTE
-	pgste = *(unsigned long *)(ptep + PTRS_PER_PTE);
-#endif
-	return __pgste(pgste);
-}
-
-static inline void pgste_set(pte_t *ptep, pgste_t pgste)
-{
-#ifdef CONFIG_PGSTE
-	*(pgste_t *)(ptep + PTRS_PER_PTE) = pgste;
-#endif
-}
-
-static inline pgste_t pgste_update_all(pte_t *ptep, pgste_t pgste,
-				       struct mm_struct *mm)
-{
-#ifdef CONFIG_PGSTE
-	unsigned long address, bits, skey;
-
-	if (!mm_use_skey(mm) || pte_val(*ptep) & _PAGE_INVALID)
-		return pgste;
-	address = pte_val(*ptep) & PAGE_MASK;
-	skey = (unsigned long) page_get_storage_key(address);
-	bits = skey & (_PAGE_CHANGED | _PAGE_REFERENCED);
-	/* Transfer page changed & referenced bit to guest bits in pgste */
-	pgste_val(pgste) |= bits << 48;		/* GR bit & GC bit */
-	/* Copy page access key and fetch protection bit to pgste */
-	pgste_val(pgste) &= ~(PGSTE_ACC_BITS | PGSTE_FP_BIT);
-	pgste_val(pgste) |= (skey & (_PAGE_ACC_BITS | _PAGE_FP_BIT)) << 56;
-#endif
-	return pgste;
-
-}
-
-static inline void pgste_set_key(pte_t *ptep, pgste_t pgste, pte_t entry,
-				 struct mm_struct *mm)
-{
-#ifdef CONFIG_PGSTE
-	unsigned long address;
-	unsigned long nkey;
-
-	if (!mm_use_skey(mm) || pte_val(entry) & _PAGE_INVALID)
-		return;
-	VM_BUG_ON(!(pte_val(*ptep) & _PAGE_INVALID));
-	address = pte_val(entry) & PAGE_MASK;
-	/*
-	 * Set page access key and fetch protection bit from pgste.
-	 * The guest C/R information is still in the PGSTE, set real
-	 * key C/R to 0.
-	 */
-	nkey = (pgste_val(pgste) & (PGSTE_ACC_BITS | PGSTE_FP_BIT)) >> 56;
-	nkey |= (pgste_val(pgste) & (PGSTE_GR_BIT | PGSTE_GC_BIT)) >> 48;
-	page_set_storage_key(address, nkey, 0);
-#endif
-}
-
-static inline pgste_t pgste_set_pte(pte_t *ptep, pgste_t pgste, pte_t entry)
-{
-	if ((pte_val(entry) & _PAGE_PRESENT) &&
-	    (pte_val(entry) & _PAGE_WRITE) &&
-	    !(pte_val(entry) & _PAGE_INVALID)) {
-		if (!MACHINE_HAS_ESOP) {
-			/*
-			 * Without enhanced suppression-on-protection force
-			 * the dirty bit on for all writable ptes.
-			 */
-			pte_val(entry) |= _PAGE_DIRTY;
-			pte_val(entry) &= ~_PAGE_PROTECT;
-		}
-		if (!(pte_val(entry) & _PAGE_PROTECT))
-			/* This pte allows write access, set user-dirty */
-			pgste_val(pgste) |= PGSTE_UC_BIT;
-	}
-	*ptep = entry;
-	return pgste;
-}
-
-/**
- * struct gmap_struct - guest address space
- * @crst_list: list of all crst tables used in the guest address space
- * @mm: pointer to the parent mm_struct
- * @guest_to_host: radix tree with guest to host address translation
- * @host_to_guest: radix tree with pointer to segment table entries
- * @guest_table_lock: spinlock to protect all entries in the guest page table
- * @table: pointer to the page directory
- * @asce: address space control element for gmap page table
- * @pfault_enabled: defines if pfaults are applicable for the guest
- */
-struct gmap {
-	struct list_head list;
-	struct list_head crst_list;
-	struct mm_struct *mm;
-	struct radix_tree_root guest_to_host;
-	struct radix_tree_root host_to_guest;
-	spinlock_t guest_table_lock;
-	unsigned long *table;
-	unsigned long asce;
-	unsigned long asce_end;
-	void *private;
-	bool pfault_enabled;
-};
-
-/**
- * struct gmap_notifier - notify function block for page invalidation
- * @notifier_call: address of callback function
- */
-struct gmap_notifier {
-	struct list_head list;
-	void (*notifier_call)(struct gmap *gmap, unsigned long gaddr);
-};
-
-struct gmap *gmap_alloc(struct mm_struct *mm, unsigned long limit);
-void gmap_free(struct gmap *gmap);
-void gmap_enable(struct gmap *gmap);
-void gmap_disable(struct gmap *gmap);
-int gmap_map_segment(struct gmap *gmap, unsigned long from,
-		     unsigned long to, unsigned long len);
-int gmap_unmap_segment(struct gmap *gmap, unsigned long to, unsigned long len);
-unsigned long __gmap_translate(struct gmap *, unsigned long gaddr);
-unsigned long gmap_translate(struct gmap *, unsigned long gaddr);
-int __gmap_link(struct gmap *gmap, unsigned long gaddr, unsigned long vmaddr);
-int gmap_fault(struct gmap *, unsigned long gaddr, unsigned int fault_flags);
-void gmap_discard(struct gmap *, unsigned long from, unsigned long to);
-void __gmap_zap(struct gmap *, unsigned long gaddr);
-bool gmap_test_and_clear_dirty(unsigned long address, struct gmap *);
-
-
-void gmap_register_ipte_notifier(struct gmap_notifier *);
-void gmap_unregister_ipte_notifier(struct gmap_notifier *);
-int gmap_ipte_notify(struct gmap *, unsigned long start, unsigned long len);
-void gmap_do_ipte_notify(struct mm_struct *, unsigned long addr, pte_t *);
-
-static inline pgste_t pgste_ipte_notify(struct mm_struct *mm,
-					unsigned long addr,
-					pte_t *ptep, pgste_t pgste)
-{
-#ifdef CONFIG_PGSTE
-	if (pgste_val(pgste) & PGSTE_IN_BIT) {
-		pgste_val(pgste) &= ~PGSTE_IN_BIT;
-		gmap_do_ipte_notify(mm, addr, ptep);
-	}
-#endif
-	return pgste;
-}
-
-/*
- * Certain architectures need to do special things when PTEs
- * within a page table are directly modified.  Thus, the following
- * hook is made available.
- */
-static inline void set_pte_at(struct mm_struct *mm, unsigned long addr,
-			      pte_t *ptep, pte_t entry)
-{
-	pgste_t pgste;
-
-	if (mm_has_pgste(mm)) {
-		pgste = pgste_get_lock(ptep);
-		pgste_val(pgste) &= ~_PGSTE_GPS_ZERO;
-		pgste_set_key(ptep, pgste, entry, mm);
-		pgste = pgste_set_pte(ptep, pgste, entry);
-		pgste_set_unlock(ptep, pgste);
-	} else {
-		*ptep = entry;
-	}
 }
 
 /*
···
 	} while (nr != 255);
 }
 
-static inline void ptep_flush_direct(struct mm_struct *mm,
-				     unsigned long address, pte_t *ptep)
-{
-	int active, count;
-
-	if (pte_val(*ptep) & _PAGE_INVALID)
-		return;
-	active = (mm == current->active_mm) ? 1 : 0;
-	count = atomic_add_return(0x10000, &mm->context.attach_count);
-	if (MACHINE_HAS_TLB_LC && (count & 0xffff) <= active &&
-	    cpumask_equal(mm_cpumask(mm), cpumask_of(smp_processor_id())))
-		__ptep_ipte_local(address, ptep);
-	else
-		__ptep_ipte(address, ptep);
-	atomic_sub(0x10000, &mm->context.attach_count);
-}
-
-static inline void ptep_flush_lazy(struct mm_struct *mm,
-				   unsigned long address, pte_t *ptep)
-{
-	int active, count;
-
-	if (pte_val(*ptep) & _PAGE_INVALID)
-		return;
-	active = (mm == current->active_mm) ? 1 : 0;
-	count = atomic_add_return(0x10000, &mm->context.attach_count);
-	if ((count & 0xffff) <= active) {
-		pte_val(*ptep) |= _PAGE_INVALID;
-		mm->context.flush_mm = 1;
-	} else
-		__ptep_ipte(address, ptep);
-	atomic_sub(0x10000, &mm->context.attach_count);
-}
-
-/*
- * Get (and clear) the user dirty bit for a pte.
- */
-static inline int ptep_test_and_clear_user_dirty(struct mm_struct *mm,
-						 unsigned long addr,
-						 pte_t *ptep)
-{
-	pgste_t pgste;
-	pte_t pte;
-	int dirty;
-
-	if (!mm_has_pgste(mm))
-		return 0;
-	pgste = pgste_get_lock(ptep);
-	dirty = !!(pgste_val(pgste) & PGSTE_UC_BIT);
-	pgste_val(pgste) &= ~PGSTE_UC_BIT;
-	pte = *ptep;
-	if (dirty && (pte_val(pte) & _PAGE_PRESENT)) {
-		pgste = pgste_ipte_notify(mm, addr, ptep, pgste);
-		__ptep_ipte(addr, ptep);
-		if (MACHINE_HAS_ESOP || !(pte_val(pte) & _PAGE_WRITE))
-			pte_val(pte) |= _PAGE_PROTECT;
-		else
-			pte_val(pte) |= _PAGE_INVALID;
-		*ptep = pte;
-	}
-	pgste_set_unlock(ptep, pgste);
-	return dirty;
-}
-
-#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
-static inline int ptep_test_and_clear_young(struct vm_area_struct *vma,
-					    unsigned long addr, pte_t *ptep)
-{
-	pgste_t pgste;
-	pte_t pte, oldpte;
-	int young;
-
-	if (mm_has_pgste(vma->vm_mm)) {
-		pgste = pgste_get_lock(ptep);
-		pgste = pgste_ipte_notify(vma->vm_mm, addr, ptep, pgste);
-	}
-
-	oldpte = pte = *ptep;
-	ptep_flush_direct(vma->vm_mm, addr, ptep);
-	young = pte_young(pte);
-	pte = pte_mkold(pte);
-
-	if (mm_has_pgste(vma->vm_mm)) {
-		pgste = pgste_update_all(&oldpte, pgste, vma->vm_mm);
-		pgste = pgste_set_pte(ptep, pgste, pte);
-		pgste_set_unlock(ptep, pgste);
-	} else
-		*ptep = pte;
-
-	return young;
-}
-
-#define __HAVE_ARCH_PTEP_CLEAR_YOUNG_FLUSH
-static inline int ptep_clear_flush_young(struct vm_area_struct *vma,
-					 unsigned long address, pte_t *ptep)
-{
-	return ptep_test_and_clear_young(vma, address, ptep);
-}
-
 /*
  * This is hard to understand. ptep_get_and_clear and ptep_clear_flush
  * both clear the TLB for the unmapped pte. The reason is that
···
  * have ptep_get_and_clear do the tlb flush. In exchange flush_tlb_range
  * is a nop.
  */
+pte_t ptep_xchg_direct(struct mm_struct *, unsigned long, pte_t *, pte_t);
+pte_t ptep_xchg_lazy(struct mm_struct *, unsigned long, pte_t *, pte_t);
+
+#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
+static inline int ptep_test_and_clear_young(struct vm_area_struct *vma,
+					    unsigned long addr, pte_t *ptep)
+{
+	pte_t pte = *ptep;
+
+	pte = ptep_xchg_direct(vma->vm_mm, addr, ptep, pte_mkold(pte));
+	return pte_young(pte);
+}
+
+#define __HAVE_ARCH_PTEP_CLEAR_YOUNG_FLUSH
+static inline int ptep_clear_flush_young(struct vm_area_struct *vma,
+					 unsigned long address, pte_t *ptep)
+{
+	return ptep_test_and_clear_young(vma, address, ptep);
+}
+
 #define __HAVE_ARCH_PTEP_GET_AND_CLEAR
 static inline pte_t ptep_get_and_clear(struct mm_struct *mm,
-				       unsigned long address, pte_t *ptep)
+				       unsigned long addr, pte_t *ptep)
 {
-	pgste_t pgste;
-	pte_t pte;
-
-	if (mm_has_pgste(mm)) {
-		pgste = pgste_get_lock(ptep);
-		pgste = pgste_ipte_notify(mm, address, ptep, pgste);
-	}
-
-	pte = *ptep;
-	ptep_flush_lazy(mm, address, ptep);
-	pte_val(*ptep) = _PAGE_INVALID;
-
-	if (mm_has_pgste(mm)) {
-		pgste = pgste_update_all(&pte, pgste, mm);
-		pgste_set_unlock(ptep, pgste);
-	}
-	return pte;
+	return ptep_xchg_lazy(mm, addr, ptep, __pte(_PAGE_INVALID));
 }
 
 #define __HAVE_ARCH_PTEP_MODIFY_PROT_TRANSACTION
-static inline pte_t ptep_modify_prot_start(struct mm_struct *mm,
-					   unsigned long address,
-					   pte_t *ptep)
-{
-	pgste_t pgste;
-	pte_t pte;
-
-	if (mm_has_pgste(mm)) {
-		pgste = pgste_get_lock(ptep);
-		pgste_ipte_notify(mm, address, ptep, pgste);
-	}
-
-	pte = *ptep;
-	ptep_flush_lazy(mm, address, ptep);
-
-	if (mm_has_pgste(mm)) {
-		pgste = pgste_update_all(&pte, pgste, mm);
-		pgste_set(ptep, pgste);
-	}
-	return pte;
-}
-
-static inline void ptep_modify_prot_commit(struct mm_struct *mm,
-					   unsigned long address,
-					   pte_t *ptep, pte_t pte)
-{
-	pgste_t pgste;
-
-	if (mm_has_pgste(mm)) {
-		pgste = pgste_get(ptep);
-		pgste_set_key(ptep, pgste, pte, mm);
-		pgste = pgste_set_pte(ptep, pgste, pte);
-		pgste_set_unlock(ptep, pgste);
-	} else
-		*ptep = pte;
-}
+pte_t ptep_modify_prot_start(struct mm_struct *, unsigned long, pte_t *);
+void ptep_modify_prot_commit(struct mm_struct *, unsigned long, pte_t *, pte_t);
 
 #define __HAVE_ARCH_PTEP_CLEAR_FLUSH
 static inline pte_t ptep_clear_flush(struct vm_area_struct *vma,
-				     unsigned long address, pte_t *ptep)
+				     unsigned long addr, pte_t *ptep)
 {
-	pgste_t pgste;
-	pte_t pte;
-
-	if (mm_has_pgste(vma->vm_mm)) {
-		pgste = pgste_get_lock(ptep);
-		pgste = pgste_ipte_notify(vma->vm_mm, address, ptep, pgste);
-	}
-
-	pte = *ptep;
-	ptep_flush_direct(vma->vm_mm, address, ptep);
-	pte_val(*ptep) = _PAGE_INVALID;
-
-	if (mm_has_pgste(vma->vm_mm)) {
-		if ((pgste_val(pgste) & _PGSTE_GPS_USAGE_MASK) ==
-		    _PGSTE_GPS_USAGE_UNUSED)
-			pte_val(pte) |= _PAGE_UNUSED;
-		pgste = pgste_update_all(&pte, pgste, vma->vm_mm);
-		pgste_set_unlock(ptep, pgste);
-	}
-	return pte;
+	return ptep_xchg_direct(vma->vm_mm, addr, ptep, __pte(_PAGE_INVALID));
 }
 
 /*
···
  */
 #define __HAVE_ARCH_PTEP_GET_AND_CLEAR_FULL
 static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm,
-					    unsigned long address,
+					    unsigned long addr,
 					    pte_t *ptep, int full)
 {
-	pgste_t pgste;
-	pte_t pte;
-
-	if (!full && mm_has_pgste(mm)) {
-		pgste = pgste_get_lock(ptep);
-		pgste = pgste_ipte_notify(mm, address, ptep, pgste);
+	if (full) {
+		pte_t pte = *ptep;
+		*ptep = __pte(_PAGE_INVALID);
+		return pte;
 	}
-
-	pte = *ptep;
-	if (!full)
-		ptep_flush_lazy(mm, address, ptep);
-	pte_val(*ptep) = _PAGE_INVALID;
-
-	if (!full && mm_has_pgste(mm)) {
-		pgste = pgste_update_all(&pte, pgste, mm);
-		pgste_set_unlock(ptep, pgste);
-	}
-	return pte;
+	return ptep_xchg_lazy(mm, addr, ptep, __pte(_PAGE_INVALID));
 }
 
 #define __HAVE_ARCH_PTEP_SET_WRPROTECT
-static inline pte_t ptep_set_wrprotect(struct mm_struct *mm,
-				       unsigned long address, pte_t *ptep)
+static inline void ptep_set_wrprotect(struct mm_struct *mm,
+				      unsigned long addr, pte_t *ptep)
 {
-	pgste_t pgste;
 	pte_t pte = *ptep;
 
-	if (pte_write(pte)) {
-		if (mm_has_pgste(mm)) {
-			pgste = pgste_get_lock(ptep);
-			pgste = pgste_ipte_notify(mm, address, ptep, pgste);
-		}
-
-		ptep_flush_lazy(mm, address, ptep);
-		pte = pte_wrprotect(pte);
-
-		if (mm_has_pgste(mm)) {
-			pgste = pgste_set_pte(ptep, pgste, pte);
-			pgste_set_unlock(ptep, pgste);
-		} else
-			*ptep = pte;
-	}
-	return pte;
+	if (pte_write(pte))
+		ptep_xchg_lazy(mm, addr, ptep, pte_wrprotect(pte));
 }
 
 #define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
 static inline int ptep_set_access_flags(struct vm_area_struct *vma,
-					unsigned long address, pte_t *ptep,
+					unsigned long addr, pte_t *ptep,
 					pte_t entry, int dirty)
 {
-	pgste_t pgste;
-	pte_t oldpte;
-
-	oldpte = *ptep;
-	if (pte_same(oldpte, entry))
+	if (pte_same(*ptep, entry))
 		return 0;
-	if (mm_has_pgste(vma->vm_mm)) {
-		pgste = pgste_get_lock(ptep);
-		pgste = pgste_ipte_notify(vma->vm_mm, address, ptep, pgste);
-	}
-
-	ptep_flush_direct(vma->vm_mm, address, ptep);
-
-	if (mm_has_pgste(vma->vm_mm)) {
-		if (pte_val(oldpte) & _PAGE_INVALID)
-			pgste_set_key(ptep, pgste, entry, vma->vm_mm);
-		pgste = pgste_set_pte(ptep, pgste, entry);
-		pgste_set_unlock(ptep, pgste);
-	} else
-		*ptep = entry;
+	ptep_xchg_direct(vma->vm_mm, addr, ptep, entry);
 	return 1;
+}
+
+/*
+ * Additional functions to handle KVM guest page tables
+ */
+void ptep_set_pte_at(struct mm_struct *mm, unsigned long addr,
+		     pte_t *ptep, pte_t entry);
+void ptep_set_notify(struct mm_struct *mm, unsigned long addr, pte_t *ptep);
+void ptep_notify(struct mm_struct *mm, unsigned long addr, pte_t *ptep);
+void ptep_zap_unused(struct mm_struct *mm, unsigned long addr,
+		     pte_t *ptep , int reset);
+void ptep_zap_key(struct mm_struct *mm, unsigned long addr, pte_t *ptep);
+
+bool test_and_clear_guest_dirty(struct mm_struct *mm, unsigned long address);
+int set_guest_storage_key(struct mm_struct *mm, unsigned long addr,
+			  unsigned char key, bool nq);
+unsigned char get_guest_storage_key(struct mm_struct *mm, unsigned long addr);
+
+/*
+ * Certain architectures need to do special things when PTEs
+ * within a page table are directly modified.  Thus, the following
+ * hook is made available.
+ */
+static inline void set_pte_at(struct mm_struct *mm, unsigned long addr,
+			      pte_t *ptep, pte_t entry)
+{
+	if (mm_has_pgste(mm))
+		ptep_set_pte_at(mm, addr, ptep, entry);
+	else
+		*ptep = entry;
 }
 
 /*
···
 		: "cc" );
 }
 
-static inline void pmdp_flush_direct(struct mm_struct *mm,
-				     unsigned long address, pmd_t *pmdp)
-{
-	int active, count;
-
-	if (pmd_val(*pmdp) & _SEGMENT_ENTRY_INVALID)
-		return;
-	if (!MACHINE_HAS_IDTE) {
-		__pmdp_csp(pmdp);
-		return;
-	}
-	active = (mm == current->active_mm) ? 1 : 0;
-	count = atomic_add_return(0x10000, &mm->context.attach_count);
-	if (MACHINE_HAS_TLB_LC && (count & 0xffff) <= active &&
-	    cpumask_equal(mm_cpumask(mm), cpumask_of(smp_processor_id())))
-		__pmdp_idte_local(address, pmdp);
-	else
-		__pmdp_idte(address, pmdp);
-	atomic_sub(0x10000, &mm->context.attach_count);
-}
-
-static inline void pmdp_flush_lazy(struct mm_struct *mm,
-				   unsigned long address, pmd_t *pmdp)
-{
-	int active, count;
-
-	if (pmd_val(*pmdp) & _SEGMENT_ENTRY_INVALID)
-		return;
-	active = (mm == current->active_mm) ? 1 : 0;
-	count = atomic_add_return(0x10000, &mm->context.attach_count);
-	if ((count & 0xffff) <= active) {
-		pmd_val(*pmdp) |= _SEGMENT_ENTRY_INVALID;
-		mm->context.flush_mm = 1;
-	} else if (MACHINE_HAS_IDTE)
-		__pmdp_idte(address, pmdp);
-	else
-		__pmdp_csp(pmdp);
-	atomic_sub(0x10000, &mm->context.attach_count);
-}
+pmd_t pmdp_xchg_direct(struct mm_struct *, unsigned long, pmd_t *, pmd_t);
+pmd_t pmdp_xchg_lazy(struct mm_struct *, unsigned long, pmd_t *, pmd_t);
 
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
 
 #define __HAVE_ARCH_PGTABLE_DEPOSIT
-extern void pgtable_trans_huge_deposit(struct mm_struct *mm, pmd_t *pmdp,
-				       pgtable_t pgtable);
+void pgtable_trans_huge_deposit(struct mm_struct *mm, pmd_t *pmdp,
+				pgtable_t pgtable);
 
 #define __HAVE_ARCH_PGTABLE_WITHDRAW
-extern pgtable_t pgtable_trans_huge_withdraw(struct mm_struct *mm, pmd_t *pmdp);
+pgtable_t pgtable_trans_huge_withdraw(struct mm_struct *mm, pmd_t *pmdp);
+
+#define  __HAVE_ARCH_PMDP_SET_ACCESS_FLAGS
+static inline int pmdp_set_access_flags(struct vm_area_struct *vma,
+					unsigned long addr, pmd_t *pmdp,
+					pmd_t entry, int dirty)
+{
+	VM_BUG_ON(addr & ~HPAGE_MASK);
+
+	entry = pmd_mkyoung(entry);
+	if (dirty)
+		entry = pmd_mkdirty(entry);
+	if (pmd_val(*pmdp) == pmd_val(entry))
+		return 0;
+	pmdp_xchg_direct(vma->vm_mm, addr, pmdp, entry);
+	return 1;
+}
+
+#define __HAVE_ARCH_PMDP_TEST_AND_CLEAR_YOUNG
+static inline int pmdp_test_and_clear_young(struct vm_area_struct *vma,
+					    unsigned long addr, pmd_t *pmdp)
+{
+	pmd_t pmd = *pmdp;
+
+	pmd = pmdp_xchg_direct(vma->vm_mm, addr, pmdp, pmd_mkold(pmd));
+	return pmd_young(pmd);
+}
+
+#define __HAVE_ARCH_PMDP_CLEAR_YOUNG_FLUSH
+static inline int pmdp_clear_flush_young(struct vm_area_struct *vma,
+					 unsigned long addr, pmd_t *pmdp)
+{
+	VM_BUG_ON(addr & ~HPAGE_MASK);
+	return pmdp_test_and_clear_young(vma, addr, pmdp);
+}
 
 static inline void set_pmd_at(struct mm_struct *mm, unsigned long addr,
 			      pmd_t *pmdp, pmd_t entry)
···
 	return pmd;
 }
 
-#define __HAVE_ARCH_PMDP_TEST_AND_CLEAR_YOUNG
-static inline int pmdp_test_and_clear_young(struct vm_area_struct *vma,
-					    unsigned long address, pmd_t *pmdp)
-{
-	pmd_t pmd;
-
-	pmd = *pmdp;
-	pmdp_flush_direct(vma->vm_mm, address, pmdp);
-	*pmdp = pmd_mkold(pmd);
-	return pmd_young(pmd);
-}
-
 #define __HAVE_ARCH_PMDP_HUGE_GET_AND_CLEAR
 static inline pmd_t pmdp_huge_get_and_clear(struct mm_struct *mm,
-					    unsigned long address, pmd_t *pmdp)
+					    unsigned long addr, pmd_t *pmdp)
 {
-	pmd_t pmd = *pmdp;
-
-	pmdp_flush_direct(mm, address, pmdp);
-	pmd_clear(pmdp);
-	return pmd;
+	return pmdp_xchg_direct(mm, addr, pmdp, __pmd(_SEGMENT_ENTRY_INVALID));
 }
 
 #define __HAVE_ARCH_PMDP_HUGE_GET_AND_CLEAR_FULL
 static inline pmd_t pmdp_huge_get_and_clear_full(struct mm_struct *mm,
-						 unsigned long address,
+						 unsigned long addr,
 						 pmd_t *pmdp, int full)
 {
-	pmd_t pmd = *pmdp;
-
-	if (!full)
-		pmdp_flush_lazy(mm, address, pmdp);
-	pmd_clear(pmdp);
-	return pmd;
+	if (full) {
+		pmd_t pmd = *pmdp;
+		*pmdp = __pmd(_SEGMENT_ENTRY_INVALID);
+		return pmd;
+	}
+	return pmdp_xchg_lazy(mm, addr, pmdp, __pmd(_SEGMENT_ENTRY_INVALID));
 }
 
 #define __HAVE_ARCH_PMDP_HUGE_CLEAR_FLUSH
 static inline pmd_t pmdp_huge_clear_flush(struct vm_area_struct *vma,
-					  unsigned long address, pmd_t *pmdp)
+					  unsigned long addr, pmd_t *pmdp)
 {
-	return pmdp_huge_get_and_clear(vma->vm_mm, address, pmdp);
+	return pmdp_huge_get_and_clear(vma->vm_mm, addr, pmdp);
 }
 
 #define __HAVE_ARCH_PMDP_INVALIDATE
 static inline void pmdp_invalidate(struct vm_area_struct *vma,
-				   unsigned long address, pmd_t *pmdp)
+				   unsigned long addr, pmd_t *pmdp)
 {
-	pmdp_flush_direct(vma->vm_mm, address, pmdp);
+	pmdp_xchg_direct(vma->vm_mm, addr, pmdp, __pmd(_SEGMENT_ENTRY_INVALID));
 }
 
 #define __HAVE_ARCH_PMDP_SET_WRPROTECT
 static inline void pmdp_set_wrprotect(struct mm_struct *mm,
-				      unsigned long address, pmd_t *pmdp)
+				      unsigned long addr, pmd_t *pmdp)
 {
 	pmd_t pmd = *pmdp;
 
-	if (pmd_write(pmd)) {
-		pmdp_flush_direct(mm, address, pmdp);
-		set_pmd_at(mm, address, pmdp, pmd_wrprotect(pmd));
-	}
+	if (pmd_write(pmd))
+		pmd = pmdp_xchg_lazy(mm, addr, pmdp, pmd_wrprotect(pmd));
 }
 
 static inline pmd_t pmdp_collapse_flush(struct vm_area_struct *vma,

arch/s390/include/asm/processor.h

···
 struct mm_struct;
 struct seq_file;
 
+typedef int (*dump_trace_func_t)(void *data, unsigned long address);
+void dump_trace(dump_trace_func_t func, void *data,
+		struct task_struct *task, unsigned long sp);
+
 void show_cacheinfo(struct seq_file *m);
 
 /* Free all resources held by a thread. */
···
 
 /* Has task runtime instrumentation enabled ? */
 #define is_ri_task(tsk) (!!(tsk)->thread.ri_cb)
+
+static inline unsigned long current_stack_pointer(void)
+{
+	unsigned long sp;
+
+	asm volatile("la %0,0(15)" : "=a" (sp));
+	return sp;
+}
 
 static inline unsigned short stap(void)
 {

arch/s390/include/asm/rwsem.h

···
  * This should be totally fair - if anything is waiting, a process that wants a
  * lock will go to the back of the queue. When the currently active lock is
  * released, if there's a writer at the front of the queue, then that and only
- * that will be woken up; if there's a bunch of consequtive readers at the
+ * that will be woken up; if there's a bunch of consecutive readers at the
  * front, then they'll all be woken up, but no other readers will be.
  */
 

arch/s390/include/asm/setup.h

···
 #define pfault_fini()		do { } while (0)
 #endif /* CONFIG_PFAULT */
 
+void report_user_fault(struct pt_regs *regs, long signr, int is_mm_fault);
+
 extern void cmma_init(void);
 
 extern void (*_machine_restart)(char *command);

arch/s390/include/asm/xor.h

···
-#include <asm-generic/xor.h>
+/*
+ * Optimited xor routines
+ *
+ * Copyright IBM Corp. 2016
+ * Author(s): Martin Schwidefsky <[e-mail]>
+ */
+#ifndef _ASM_S390_XOR_H
+#define _ASM_S390_XOR_H
+
+extern struct xor_block_template xor_block_xc;
+
+#undef XOR_TRY_TEMPLATES
+#define XOR_TRY_TEMPLATES				\
+do {							\
+	xor_speed(&xor_block_xc);			\
+} while (0)
+
+#define XOR_SELECT_TEMPLATE(FASTEST)	(&xor_block_xc)
+
+#endif /* _ASM_S390_XOR_H */

arch/s390/include/uapi/asm/clp.h

···
+/*
+ * ioctl interface for /dev/clp
+ *
+ * Copyright IBM Corp. 2016
+ * Author(s): Martin Schwidefsky <[e-mail]>
+ */
+
+#ifndef _ASM_CLP_H
+#define _ASM_CLP_H
+
+#include <linux/types.h>
+#include <linux/ioctl.h>
+
+struct clp_req {
+	unsigned int c : 1;
+	unsigned int r : 1;
+	unsigned int lps : 6;
+	unsigned int cmd : 8;
+	unsigned int : 16;
+	unsigned int reserved;
+	__u64 data_p;
+};
+
+#define CLP_IOCTL_MAGIC 'c'
+
+#define CLP_SYNC _IOWR(CLP_IOCTL_MAGIC, 0xC1, struct clp_req)
+
+#endif

arch/s390/kernel/asm-offsets.c

···
 #include <asm/idle.h>
 #include <asm/vdso.h>
 #include <asm/pgtable.h>
+#include <asm/gmap.h>
 
 /*
  * Make sure that the compiler is new enough. We want a compiler that

arch/s390/kernel/cpcmd.c

···
 			(((unsigned long)response + rlen) >> 31)) {
 		lowbuf = kmalloc(rlen, GFP_KERNEL | GFP_DMA);
 		if (!lowbuf) {
-			pr_warning("The cpcmd kernel function failed to "
-				   "allocate a response buffer\n");
+			pr_warn("The cpcmd kernel function failed to allocate a response buffer\n");
 			return -ENOMEM;
 		}
 		spin_lock_irqsave(&cpcmd_lock, flags);

arch/s390/kernel/debug.c

···
 	/* Since debugfs currently does not support uid/gid other than root, */
 	/* we do not allow gid/uid != 0 until we get support for that. */
 	if ((uid != 0) || (gid != 0))
-		pr_warning("Root becomes the owner of all s390dbf files "
-			   "in sysfs\n");
+		pr_warn("Root becomes the owner of all s390dbf files in sysfs\n");
 	BUG_ON(!initialized);
 	mutex_lock(&debug_mutex);
 
···
 		new_level = debug_get_uint(str);
 	}
 	if(new_level < 0) {
-		pr_warning("%s is not a valid level for a debug "
-			   "feature\n", str);
+		pr_warn("%s is not a valid level for a debug feature\n", str);
 		rc = -EINVAL;
 	} else {
 		debug_set_level(id, new_level);

arch/s390/kernel/dis.c

···
 			}
 			if (separator)
 				ptr += sprintf(ptr, "%c", separator);
-			/*
-			 * Use four '%' characters below because of the
-			 * following two conversions:
-			 *
-			 *  1) sprintf: %%%%r -> %%r
-			 *  2) printk : %%r   -> %r
-			 */
 			if (operand->flags & OPERAND_GPR)
-				ptr += sprintf(ptr, "%%%%r%i", value);
+				ptr += sprintf(ptr, "%%r%i", value);
 			else if (operand->flags & OPERAND_FPR)
-				ptr += sprintf(ptr, "%%%%f%i", value);
+				ptr += sprintf(ptr, "%%f%i", value);
 			else if (operand->flags & OPERAND_AR)
-				ptr += sprintf(ptr, "%%%%a%i", value);
+				ptr += sprintf(ptr, "%%a%i", value);
 			else if (operand->flags & OPERAND_CR)
-				ptr += sprintf(ptr, "%%%%c%i", value);
+				ptr += sprintf(ptr, "%%c%i", value);
 			else if (operand->flags & OPERAND_VR)
-				ptr += sprintf(ptr, "%%%%v%i", value);
+				ptr += sprintf(ptr, "%%v%i", value);
 			else if (operand->flags & OPERAND_PCREL)
 				ptr += sprintf(ptr, "%lx", (signed int) value
 								      + addr);

arch/s390/kernel/dumpstack.c

···
 #include <asm/ipl.h>
 
 /*
- * For show_trace we have tree different stack to consider:
+ * For dump_trace we have tree different stack to consider:
  *   - the panic stack which is used if the kernel stack has overflown
  *   - the asynchronous interrupt stack (cpu related)
  *   - the synchronous kernel stack (process related)
- * The stack trace can start at any of the three stack and can potentially
+ * The stack trace can start at any of the three stacks and can potentially
  * touch all of them. The order is: panic stack, async stack, sync stack.
  */
 static unsigned long
-__show_trace(unsigned long sp, unsigned long low, unsigned long high)
+__dump_trace(dump_trace_func_t func, void *data, unsigned long sp,
+	     unsigned long low, unsigned long high)
 {
 	struct stack_frame *sf;
 	struct pt_regs *regs;
-	unsigned long addr;
 
 	while (1) {
 		if (sp < low || sp > high - sizeof(*sf))
 			return sp;
 		sf = (struct stack_frame *) sp;
-		addr = sf->gprs[8];
-		printk("([<%016lx>] %pSR)\n", addr, (void *)addr);
 		/* Follow the backchain. */
 		while (1) {
+			if (func(data, sf->gprs[8]))
+				return sp;
 			low = sp;
 			sp = sf->back_chain;
 			if (!sp)
···
 			if (sp <= low || sp > high - sizeof(*sf))
 				return sp;
 			sf = (struct stack_frame *) sp;
-			addr = sf->gprs[8];
-			printk(" [<%016lx>] %pSR\n", addr, (void *)addr);
 		}
 		/* Zero backchain detected, check for interrupt frame. */
 		sp = (unsigned long) (sf + 1);
 		if (sp <= low || sp > high - sizeof(*regs))
 			return sp;
 		regs = (struct pt_regs *) sp;
-		addr = regs->psw.addr;
-		printk(" [<%016lx>] %pSR\n", addr, (void *)addr);
+		if (!user_mode(regs)) {
+			if (func(data, regs->psw.addr))
+				return sp;
+		}
 		low = sp;
 		sp = regs->gprs[15];
 	}
 }
 
-static void show_trace(struct task_struct *task, unsigned long *stack)
+void dump_trace(dump_trace_func_t func, void *data, struct task_struct *task,
+		unsigned long sp)
 {
-	const unsigned long frame_size =
-		STACK_FRAME_OVERHEAD + sizeof(struct pt_regs);
-	register unsigned long __r15 asm ("15");
-	unsigned long sp;
+	unsigned long frame_size;
 
-	sp = (unsigned long) stack;
-	if (!sp)
-		sp = task ? task->thread.ksp : __r15;
-	printk("Call Trace:\n");
+	frame_size = STACK_FRAME_OVERHEAD + sizeof(struct pt_regs);
 #ifdef CONFIG_CHECK_STACK
-	sp = __show_trace(sp,
+	sp = __dump_trace(func, data, sp,
 			  S390_lowcore.panic_stack + frame_size - 4096,
 			  S390_lowcore.panic_stack + frame_size);
 #endif
-	sp = __show_trace(sp,
+	sp = __dump_trace(func, data, sp,
 			  S390_lowcore.async_stack + frame_size - ASYNC_SIZE,
 			  S390_lowcore.async_stack + frame_size);
 	if (task)
-		__show_trace(sp, (unsigned long) task_stack_page(task),
-			     (unsigned long) task_stack_page(task) + THREAD_SIZE);
+		__dump_trace(func, data, sp,
+			     (unsigned long)task_stack_page(task),
+			     (unsigned long)task_stack_page(task) + THREAD_SIZE);
 	else
-		__show_trace(sp, S390_lowcore.thread_info,
+		__dump_trace(func, data, sp,
+			     S390_lowcore.thread_info,
 			     S390_lowcore.thread_info + THREAD_SIZE);
+}
+EXPORT_SYMBOL_GPL(dump_trace);
+
+static int show_address(void *data, unsigned long address)
+{
+	printk("([<%016lx>] %pSR)\n", address, (void *)address);
+	return 0;
+}
+
+static void show_trace(struct task_struct *task, unsigned long sp)
+{
+	if (!sp)
+		sp = task ? task->thread.ksp : current_stack_pointer();
+	printk("Call Trace:\n");
+	dump_trace(show_address, NULL, task, sp);
 	if (!task)
 		task = current;
 	debug_show_held_locks(task);
···
 
 void show_stack(struct task_struct *task, unsigned long *sp)
 {
-	register unsigned long *__r15 asm ("15");
 	unsigned long *stack;
 	int i;
 
-	if (!sp)
-		stack = task ? (unsigned long *) task->thread.ksp : __r15;
-	else
-		stack = sp;
-
+	stack = sp;
+	if (!stack) {
+		if (!task)
+			stack = (unsigned long *)current_stack_pointer();
+		else
+			stack = (unsigned long *)task->thread.ksp;
+	}
 	for (i = 0; i < 20; i++) {
 		if (((addr_t) stack & (THREAD_SIZE-1)) == 0)
 			break;
···
 		printk("%016lx ", *stack++);
 	}
 	printk("\n");
-	show_trace(task, sp);
+	show_trace(task, (unsigned long)sp);
 }
 
 static void show_last_breaking_event(struct pt_regs *regs)
···
 	printk(" [<%016lx>] %pSR\n", regs->args[0], (void *)regs->args[0]);
 }
 
-static inline int mask_bits(struct pt_regs *regs, unsigned long bits)
-{
-	return (regs->psw.mask & bits) / ((~bits + 1) & bits);
-}
-
 void show_registers(struct pt_regs *regs)
 {
+	struct psw_bits *psw = &psw_bits(regs->psw);
 	char *mode;
 
 	mode = user_mode(regs) ? "User" : "Krnl";
···
 		printk(" (%pSR)", (void *)regs->psw.addr);
 	printk("\n");
 	printk("           R:%x T:%x IO:%x EX:%x Key:%x M:%x W:%x "
-	       "P:%x AS:%x CC:%x PM:%x", mask_bits(regs, PSW_MASK_PER),
-	       mask_bits(regs, PSW_MASK_DAT), mask_bits(regs, PSW_MASK_IO),
-	       mask_bits(regs, PSW_MASK_EXT), mask_bits(regs, PSW_MASK_KEY),
-	       mask_bits(regs, PSW_MASK_MCHECK), mask_bits(regs, PSW_MASK_WAIT),
-	       mask_bits(regs, PSW_MASK_PSTATE), mask_bits(regs, PSW_MASK_ASC),
-	       mask_bits(regs, PSW_MASK_CC), mask_bits(regs, PSW_MASK_PM));
-	printk(" EA:%x", mask_bits(regs, PSW_MASK_EA | PSW_MASK_BA));
+	       "P:%x AS:%x CC:%x PM:%x", psw->r, psw->t, psw->i, psw->e,
+	       psw->key, psw->m, psw->w, psw->p, psw->as, psw->cc, psw->pm);
+	printk(" RI:%x EA:%x", psw->ri, psw->eaba);
 	printk("\n%s GPRS: %016lx %016lx %016lx %016lx\n", mode,
 	       regs->gprs[0], regs->gprs[1], regs->gprs[2], regs->gprs[3]);
 	printk("           %016lx %016lx %016lx %016lx\n",
···
 	show_registers(regs);
 	/* Show stack backtrace if pt_regs is from kernel mode */
 	if (!user_mode(regs))
-		show_trace(NULL, (unsigned long *) regs->gprs[15]);
+		show_trace(NULL, regs->gprs[15]);
 	show_last_breaking_event(regs);
 }
 

arch/s390/kernel/entry.S

···
 	stg	%r5,__LC_THREAD_INFO		# store thread info of next
 	stg	%r15,__LC_KERNEL_STACK		# store end of kernel stack
 	lg	%r15,__THREAD_ksp(%r1)		# load kernel stack of next
+	/* c4 is used in guest detection: arch/s390/kernel/perf_cpum_sf.c */
 	lctl	%c4,%c4,__TASK_pid(%r3)		# load pid to control reg. 4
 	mvc	__LC_CURRENT_PID(4,%r0),__TASK_pid(%r3) # store pid of next
 	lmg	%r6,%r15,__SF_GPRS(%r15)	# load gprs of next task
···
 	.quad	.Lpsw_idle_lpsw
 
 .Lcleanup_save_fpu_regs:
-	TSTMSK	__LC_CPU_FLAGS,_CIF_FPU
-	bor	%r14
-	clg	%r9,BASED(.Lcleanup_save_fpu_regs_done)
-	jhe	5f
-	clg	%r9,BASED(.Lcleanup_save_fpu_regs_fp)
-	jhe	4f
-	clg	%r9,BASED(.Lcleanup_save_fpu_regs_vx_high)
-	jhe	3f
-	clg	%r9,BASED(.Lcleanup_save_fpu_regs_vx_low)
-	jhe	2f
-	clg	%r9,BASED(.Lcleanup_save_fpu_fpc_end)
-	jhe	1f
-	lg	%r2,__LC_CURRENT
-	aghi	%r2,__TASK_thread
-0:	# Store floating-point controls
-	stfpc	__THREAD_FPU_fpc(%r2)
-1:	# Load register save area and check if VX is active
-	lg	%r3,__THREAD_FPU_regs(%r2)
-	TSTMSK	__LC_MACHINE_FLAGS,MACHINE_FLAG_VX
-	jz	4f			  # no VX -> store FP regs
-2:	# Store vector registers (V0-V15)
-	VSTM	%v0,%v15,0,%r3		  # vstm 0,15,0(3)
-3:	# Store vector registers (V16-V31)
-	VSTM	%v16,%v31,256,%r3	  # vstm 16,31,256(3)
-	j	5f			  # -> done, set CIF_FPU flag
-4:	# Store floating-point registers
-	std	0,0(%r3)
-	std	1,8(%r3)
-	std	2,16(%r3)
-	std	3,24(%r3)
-	std	4,32(%r3)
-	std	5,40(%r3)
-	std	6,48(%r3)
-	std	7,56(%r3)
-	std	8,64(%r3)
-	std	9,72(%r3)
-	std	10,80(%r3)
-	std	11,88(%r3)
-	std	12,96(%r3)
-	std	13,104(%r3)
-	std	14,112(%r3)
-	std	15,120(%r3)
-5:	# Set CIF_FPU flag
-	oi	__LC_CPU_FLAGS+7,_CIF_FPU
-	lg	%r9,48(%r11)		# return from save_fpu_regs
+	larl	%r9,save_fpu_regs
 	br	%r14
-.Lcleanup_save_fpu_fpc_end:
-	.quad	.Lsave_fpu_regs_fpc_end
-.Lcleanup_save_fpu_regs_vx_low:
-	.quad	.Lsave_fpu_regs_vx_low
-.Lcleanup_save_fpu_regs_vx_high:
-	.quad	.Lsave_fpu_regs_vx_high
-.Lcleanup_save_fpu_regs_fp:
-	.quad	.Lsave_fpu_regs_fp
-.Lcleanup_save_fpu_regs_done:
-	.quad	.Lsave_fpu_regs_done
 
 .Lcleanup_load_fpu_regs:
-	TSTMSK	__LC_CPU_FLAGS,_CIF_FPU
-	bnor	%r14
-	clg	%r9,BASED(.Lcleanup_load_fpu_regs_done)
-	jhe	1f
-	clg	%r9,BASED(.Lcleanup_load_fpu_regs_fp)
-	jhe	2f
-	clg	%r9,BASED(.Lcleanup_load_fpu_regs_vx_high)
-	jhe	3f
-	clg	%r9,BASED(.Lcleanup_load_fpu_regs_vx)
-	jhe	4f
-	lg	%r4,__LC_CURRENT
-	aghi	%r4,__TASK_thread
-	lfpc	__THREAD_FPU_fpc(%r4)
-	TSTMSK	__LC_MACHINE_FLAGS,MACHINE_FLAG_VX
-	lg	%r4,__THREAD_FPU_regs(%r4)	# %r4 <- reg save area
-	jz	2f				# -> no VX, load FP regs
-4:	# Load V0 ..V15 registers
-	VLM	%v0,%v15,0,%r4
-3:	# Load V16..V31 registers
-	VLM	%v16,%v31,256,%r4
-	j	1f
-2:	# Load floating-point registers
-	ld	0,0(%r4)
-	ld	1,8(%r4)
-	ld	2,16(%r4)
-	ld	3,24(%r4)
-	ld	4,32(%r4)
-	ld	5,40(%r4)
-	ld	6,48(%r4)
-	ld	7,56(%r4)
-	ld	8,64(%r4)
-	ld	9,72(%r4)
-	ld	10,80(%r4)
-	ld	11,88(%r4)
-	ld	12,96(%r4)
-	ld	13,104(%r4)
-	ld	14,112(%r4)
-	ld	15,120(%r4)
-1:	# Clear CIF_FPU bit
-	ni	__LC_CPU_FLAGS+7,255-_CIF_FPU
-	lg	%r9,48(%r11)		# return from load_fpu_regs
+	larl	%r9,load_fpu_regs
 	br	%r14
-.Lcleanup_load_fpu_regs_vx:
-	.quad	.Lload_fpu_regs_vx
-.Lcleanup_load_fpu_regs_vx_high:
-	.quad	.Lload_fpu_regs_vx_high
-.Lcleanup_load_fpu_regs_fp:
-	.quad	.Lload_fpu_regs_fp
-.Lcleanup_load_fpu_regs_done:
-	.quad	.Lload_fpu_regs_done
 
 /*
  * Integer constants

arch/s390/kernel/irq.c

···
 {
 	unsigned long old, new;
 
-	/* Get current stack pointer. */
-	asm volatile("la %0,0(15)" : "=a" (old));
+	old = current_stack_pointer();
 	/* Check against async. stack address range. */
 	new = S390_lowcore.async_stack;
 	if (((new - old) >> (PAGE_SHIFT + THREAD_ORDER)) != 0) {

arch/s390/kernel/perf_cpum_cf.c

···
 
 	/* Validate the counter that is assigned to this event.
 	 * Because the counter facility can use numerous counters at the
-	 * same time without constraints, it is not necessary to explicity
+	 * same time without constraints, it is not necessary to explicitly
 	 * validate event groups (event->group_leader != event).
 	 */
 	err = validate_event(hwc);

arch/s390/kernel/perf_cpum_sf.c

···
 	/*
 	 * A non-zero guest program parameter indicates a guest
 	 * sample.
-	 * Note that some early samples might be misaccounted to
-	 * the host.
+	 * Note that some early samples or samples from guests without
+	 * lpp usage would be misaccounted to the host. We use the asn
+	 * value as a heuristic to detect most of these guest samples.
+	 * If the value differs from the host hpp value, we assume
+	 * it to be a KVM guest.
 	 */
-	if (sfr->basic.gpp)
+	if (sfr->basic.gpp || sfr->basic.prim_asn != (u16) sfr->basic.hpp)
 		sde_regs->in_guest = 1;
 
 	overflow = 0;

arch/s390/kernel/perf_event.c

···
 }
 arch_initcall(service_level_perf_register);
 
-/* See also arch/s390/kernel/traps.c */
-static unsigned long __store_trace(struct perf_callchain_entry *entry,
-				   unsigned long sp,
-				   unsigned long low, unsigned long high)
+static int __perf_callchain_kernel(void *data, unsigned long address)
 {
-	struct stack_frame *sf;
-	struct pt_regs *regs;
+	struct perf_callchain_entry *entry = data;
 
-	while (1) {
-		if (sp < low || sp > high - sizeof(*sf))
-			return sp;
-		sf = (struct stack_frame *) sp;
-		perf_callchain_store(entry, sf->gprs[8]);
-		/* Follow the backchain. */
-		while (1) {
-			low = sp;
-			sp = sf->back_chain;
-			if (!sp)
-				break;
-			if (sp <= low || sp > high - sizeof(*sf))
-				return sp;
-			sf = (struct stack_frame *) sp;
-			perf_callchain_store(entry, sf->gprs[8]);
-		}
-		/* Zero backchain detected, check for interrupt frame. */
-		sp = (unsigned long) (sf + 1);
-		if (sp <= low || sp > high - sizeof(*regs))
-			return sp;
-		regs = (struct pt_regs *) sp;
-		perf_callchain_store(entry, sf->gprs[8]);
-		low = sp;
-		sp = regs->gprs[15];
-	}
+	perf_callchain_store(entry, address);
+	return 0;
 }
 
 void perf_callchain_kernel(struct perf_callchain_entry *entry,
 			   struct pt_regs *regs)
 {
-	unsigned long head, frame_size;
-	struct stack_frame *head_sf;
-
 	if (user_mode(regs))
 		return;
-
-	frame_size = STACK_FRAME_OVERHEAD + sizeof(struct pt_regs);
-	head = regs->gprs[15];
-	head_sf = (struct stack_frame *) head;
-
-	if (!head_sf || !head_sf->back_chain)
-		return;
-
-	head = head_sf->back_chain;
-	head = __store_trace(entry, head,
-			     S390_lowcore.async_stack + frame_size - ASYNC_SIZE,
-			     S390_lowcore.async_stack + frame_size);
-
-	__store_trace(entry, head, S390_lowcore.thread_info,
-		      S390_lowcore.thread_info + THREAD_SIZE);
+	dump_trace(__perf_callchain_kernel, entry, NULL, regs->gprs[15]);
 }
 
-/* Perf defintions for PMU event attributes in sysfs */
+/* Perf definitions for PMU event attributes in sysfs */
 ssize_t cpumf_events_sysfs_show(struct device *dev,
 				struct device_attribute *attr, char *page)
 {

arch/s390/kernel/setup.c

···
 		+ PAGE_SIZE - STACK_FRAME_OVERHEAD - sizeof(struct pt_regs);
 	lc->current_task = (unsigned long) init_thread_union.thread_info.task;
 	lc->thread_info = (unsigned long) &init_thread_union;
+	lc->lpp = LPP_MAGIC;
 	lc->machine_flags = S390_lowcore.machine_flags;
 	lc->stfl_fac_list = S390_lowcore.stfl_fac_list;
 	memcpy(lc->stfle_fac_list, S390_lowcore.stfle_fac_list,
···
 		strcpy(elf_platform, "zEC12");
 		break;
 	case 0x2964:
+	case 0x2965:
 		strcpy(elf_platform, "z13");
 		break;
 	}

arch/s390/kernel/stacktrace.c

···
 #include <linux/kallsyms.h>
 #include <linux/module.h>
 
-static unsigned long save_context_stack(struct stack_trace *trace,
-					unsigned long sp,
-					unsigned long low,
-					unsigned long high,
-					int savesched)
+static int __save_address(void *data, unsigned long address, int nosched)
 {
-	struct stack_frame *sf;
-	struct pt_regs *regs;
-	unsigned long addr;
+	struct stack_trace *trace = data;
 
-	while(1) {
-		if (sp < low || sp > high)
-			return sp;
-		sf = (struct stack_frame *)sp;
-		while(1) {
-			addr = sf->gprs[8];
-			if (!trace->skip)
-				trace->entries[trace->nr_entries++] = addr;
-			else
-				trace->skip--;
-			if (trace->nr_entries >= trace->max_entries)
-				return sp;
-			low = sp;
-			sp = sf->back_chain;
-			if (!sp)
-				break;
-			if (sp <= low || sp > high - sizeof(*sf))
-				return sp;
-			sf = (struct stack_frame *)sp;
-		}
-		/* Zero backchain detected, check for interrupt frame. */
-		sp = (unsigned long)(sf + 1);
-		if (sp <= low || sp > high - sizeof(*regs))
-			return sp;
-		regs = (struct pt_regs *)sp;
-		addr = regs->psw.addr;
-		if (savesched || !in_sched_functions(addr)) {
-			if (!trace->skip)
-				trace->entries[trace->nr_entries++] = addr;
-			else
-				trace->skip--;
-		}
-		if (trace->nr_entries >= trace->max_entries)
-			return sp;
-		low = sp;
-		sp = regs->gprs[15];
+	if (nosched && in_sched_functions(address))
+		return 0;
+	if (trace->skip > 0) {
+		trace->skip--;
+		return 0;
 	}
+	if (trace->nr_entries < trace->max_entries) {
+		trace->entries[trace->nr_entries++] = address;
+		return 0;
+	}
+	return 1;
 }
 
-static void __save_stack_trace(struct stack_trace *trace, unsigned long sp)
+static int save_address(void *data, unsigned long address)
 {
-	unsigned long new_sp, frame_size;
+	return __save_address(data, address, 0);
+}
 
-	frame_size = STACK_FRAME_OVERHEAD + sizeof(struct pt_regs);
-	new_sp = save_context_stack(trace, sp,
-			S390_lowcore.panic_stack + frame_size - PAGE_SIZE,
-			S390_lowcore.panic_stack + frame_size, 1);
-	new_sp = save_context_stack(trace, new_sp,
-			S390_lowcore.async_stack + frame_size - ASYNC_SIZE,
-			S390_lowcore.async_stack + frame_size, 1);
-	save_context_stack(trace, new_sp,
-			   S390_lowcore.thread_info,
-			   S390_lowcore.thread_info + THREAD_SIZE, 1);
+static int save_address_nosched(void *data, unsigned long address)
+{
+	return __save_address(data, address, 1);
 }
 
 void save_stack_trace(struct stack_trace *trace)
 {
-	register unsigned long r15 asm ("15");
 	unsigned long sp;
 
-	sp = r15;
-	__save_stack_trace(trace, sp);
+	sp = current_stack_pointer();
+	dump_trace(save_address, trace, NULL, sp);
 	if (trace->nr_entries < trace->max_entries)
 		trace->entries[trace->nr_entries++] = ULONG_MAX;
 }
···
 
 void save_stack_trace_tsk(struct task_struct *tsk, struct stack_trace *trace)
 {
-	unsigned long sp, low, high;
+	unsigned long sp;
 
 	sp = tsk->thread.ksp;
-	if (tsk == current) {
-		/* Get current stack pointer. */
-		asm volatile("la %0,0(15)" : "=a" (sp));
-	}
-	low = (unsigned long) task_stack_page(tsk);
-	high = (unsigned long) task_pt_regs(tsk);
-	save_context_stack(trace, sp, low, high, 0);
+	if (tsk == current)
+		sp = current_stack_pointer();
+	dump_trace(save_address_nosched, trace, tsk, sp);
 	if (trace->nr_entries < trace->max_entries)
 		trace->entries[trace->nr_entries++] = ULONG_MAX;
 }
···
 	unsigned long sp;
 
 	sp = kernel_stack_pointer(regs);
-	__save_stack_trace(trace, sp);
+	dump_trace(save_address, trace, NULL, sp);
 	if (trace->nr_entries < trace->max_entries)
 		trace->entries[trace->nr_entries++] = ULONG_MAX;
 }

arch/s390/kernel/time.c

···
 		if (etr_port0_online && etr_port1_online)
 			set_bit(CLOCK_SYNC_ETR, &clock_sync_flags);
 	} else if (etr_port0_online || etr_port1_online) {
-		pr_warning("The real or virtual hardware system does "
-			   "not provide an ETR interface\n");
+		pr_warn("The real or virtual hardware system does not provide an ETR interface\n");
 		etr_port0_online = etr_port1_online = 0;
 	}
 }
···
 	if (rc == 0)
 		set_bit(CLOCK_SYNC_HAS_STP, &clock_sync_flags);
 	else if (stp_online) {
-		pr_warning("The real or virtual hardware system does "
-			   "not provide an STP interface\n");
+		pr_warn("The real or virtual hardware system does not provide an STP interface\n");
 		free_page((unsigned long) stp_page);
 		stp_page = NULL;
 		stp_online = 0;

arch/s390/kernel/traps.c

···
 #include <asm/fpu/api.h>
 #include "entry.h"
 
-int show_unhandled_signals = 1;
-
 static inline void __user *get_trap_ip(struct pt_regs *regs)
 {
 	unsigned long address;
···
 	else
 		address = regs->psw.addr;
 	return (void __user *) (address - (regs->int_code >> 16));
-}
-
-static inline void report_user_fault(struct pt_regs *regs, int signr)
-{
-	if ((task_pid_nr(current) > 1) && !show_unhandled_signals)
-		return;
-	if (!unhandled_signal(current, signr))
-		return;
-	if (!printk_ratelimit())
-		return;
-	printk("User process fault: interruption code %04x ilc:%d ",
-	       regs->int_code & 0xffff, regs->int_code >> 17);
-	print_vma_addr("in ", regs->psw.addr);
-	printk("\n");
-	show_regs(regs);
 }
 
 int is_valid_bugaddr(unsigned long addr)
···
 		info.si_code = si_code;
 		info.si_addr = get_trap_ip(regs);
 		force_sig_info(si_signo, &info, current);
-		report_user_fault(regs, si_signo);
+		report_user_fault(regs, si_signo, 0);
         } else {
                 const struct exception_table_entry *fixup;
 		fixup = search_exception_tables(regs->psw.addr);
···
 void default_trap_handler(struct pt_regs *regs)
 {
 	if (user_mode(regs)) {
-		report_user_fault(regs, SIGSEGV);
+		report_user_fault(regs, SIGSEGV, 0);
 		do_exit(SIGSEGV);
 	} else
 		die(regs, "Unknown program exception");

arch/s390/kvm/diag.c

···
 #include <linux/kvm.h>
 #include <linux/kvm_host.h>
 #include <asm/pgalloc.h>
+#include <asm/gmap.h>
 #include <asm/virtio-ccw.h>
 #include "kvm-s390.h"
 #include "trace.h"

arch/s390/kvm/guestdbg.c

···
 /*
  * Extends the address range given by *start and *stop to include the address
  * range starting with estart and the length len. Takes care of overflowing
- * intervals and tries to minimize the overall intervall size.
+ * intervals and tries to minimize the overall interval size.
  */
 static void extend_address_range(u64 *start, u64 *stop, u64 estart, int len)
 {
···
 		return;
 
 	/*
-	 * If the guest is not interrested in branching events, we can savely
+	 * If the guest is not interested in branching events, we can safely
 	 * limit them to the PER address range.
 	 */
 	if (!(*cr9 & PER_EVENT_BRANCH))

arch/s390/kvm/interrupt.c

···
 #include <asm/uaccess.h>
 #include <asm/sclp.h>
 #include <asm/isc.h>
+#include <asm/gmap.h>
 #include "kvm-s390.h"
 #include "gaccess.h"
 #include "trace-s390.h"

arch/s390/kvm/kvm-s390.c

···
 #include <asm/lowcore.h>
 #include <asm/etr.h>
 #include <asm/pgtable.h>
+#include <asm/gmap.h>
 #include <asm/nmi.h>
 #include <asm/switch_to.h>
 #include <asm/isc.h>
···
 	for (cur_gfn = memslot->base_gfn; cur_gfn <= last_gfn; cur_gfn++) {
 		address = gfn_to_hva_memslot(memslot, cur_gfn);
 
-		if (gmap_test_and_clear_dirty(address, gmap))
+		if (test_and_clear_guest_dirty(gmap->mm, address))
 			mark_page_dirty(kvm, cur_gfn);
 		if (fatal_signal_pending(current))
 			return;

arch/s390/kvm/priv.c

···
 #include <asm/sysinfo.h>
 #include <asm/pgtable.h>
 #include <asm/pgalloc.h>
+#include <asm/gmap.h>
 #include <asm/io.h>
 #include <asm/ptrace.h>
 #include <asm/compat.h>

arch/s390/lib/Makefile

···
 #
 
 lib-y += delay.o string.o uaccess.o find.o
-obj-y += mem.o
+obj-y += mem.o xor.o
 lib-$(CONFIG_SMP) += spinlock.o
 lib-$(CONFIG_KPROBES) += probes.o
 lib-$(CONFIG_UPROBES) += probes.o

arch/s390/lib/xor.c

···
+/*
+ * Optimized xor_block operation for RAID4/5
+ *
+ * Copyright IBM Corp. 2016
+ * Author(s): Martin Schwidefsky <[e-mail]>
+ */
+
+#include <linux/types.h>
+#include <linux/module.h>
+#include <linux/raid/xor.h>
+
+static void xor_xc_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+{
+	asm volatile(
+		"	larl	1,2f\n"
+		"	aghi	%0,-1\n"
+		"	jm	3f\n"
+		"	srlg	0,%0,8\n"
+		"	ltgr	0,0\n"
+		"	jz	1f\n"
+		"0:	xc	0(256,%1),0(%2)\n"
+		"	la	%1,256(%1)\n"
+		"	la	%2,256(%2)\n"
+		"	brctg	0,0b\n"
+		"1:	ex	%0,0(1)\n"
+		"	j	3f\n"
+		"2:	xc	0(1,%1),0(%2)\n"
+		"3:\n"
+		: : "d" (bytes), "a" (p1), "a" (p2)
+		: "0", "1", "cc", "memory");
+}
+
+static void xor_xc_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
+		     unsigned long *p3)
+{
+	asm volatile(
+		"	larl	1,2f\n"
+		"	aghi	%0,-1\n"
+		"	jm	3f\n"
+		"	srlg	0,%0,8\n"
+		"	ltgr	0,0\n"
+		"	jz	1f\n"
+		"0:	xc	0(256,%1),0(%2)\n"
+		"	xc	0(256,%1),0(%3)\n"
+		"	la	%1,256(%1)\n"
+		"	la	%2,256(%2)\n"
+		"	la	%3,256(%3)\n"
+		"	brctg	0,0b\n"
+		"1:	ex	%0,0(1)\n"
+		"	ex	%0,6(1)\n"
+		"	j	3f\n"
+		"2:	xc	0(1,%1),0(%2)\n"
+		"	xc	0(1,%1),0(%3)\n"
+		"3:\n"
+		: "+d" (bytes), "+a" (p1), "+a" (p2), "+a" (p3)
+		: : "0", "1", "cc", "memory");
+}
+
+static void xor_xc_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
+		     unsigned long *p3, unsigned long *p4)
+{
+	asm volatile(
+		"	larl	1,2f\n"
+		"	aghi	%0,-1\n"
+		"	jm	3f\n"
+		"	srlg	0,%0,8\n"
+		"	ltgr	0,0\n"
+		"	jz	1f\n"
+		"0:	xc	0(256,%1),0(%2)\n"
+		"	xc	0(256,%1),0(%3)\n"
+		"	xc	0(256,%1),0(%4)\n"
+		"	la	%1,256(%1)\n"
+		"	la	%2,256(%2)\n"
+		"	la	%3,256(%3)\n"
+		"	la	%4,256(%4)\n"
+		"	brctg	0,0b\n"
+		"1:	ex	%0,0(1)\n"
+		"	ex	%0,6(1)\n"
+		"	ex	%0,12(1)\n"
+		"	j	3f\n"
+		"2:	xc	0(1,%1),0(%2)\n"
+		"	xc	0(1,%1),0(%3)\n"
+		"	xc	0(1,%1),0(%4)\n"
+		"3:\n"
+		: "+d" (bytes), "+a" (p1), "+a" (p2), "+a" (p3), "+a" (p4)
+		: : "0", "1", "cc", "memory");
+}
+
+static void xor_xc_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
+		     unsigned long *p3, unsigned long *p4, unsigned long *p5)
+{
+	/* Get around a gcc oddity */
+	register unsigned long *reg7 asm ("7") = p5;
+
+	asm volatile(
+		"	larl	1,2f\n"
+		"	aghi	%0,-1\n"
+		"	jm	3f\n"
+		"	srlg	0,%0,8\n"
+		"	ltgr	0,0\n"
+		"	jz	1f\n"
+		"0:	xc	0(256,%1),0(%2)\n"
+		"	xc	0(256,%1),0(%3)\n"
+		"	xc	0(256,%1),0(%4)\n"
+		"	xc	0(256,%1),0(%5)\n"
+		"	la	%1,256(%1)\n"
+		"	la	%2,256(%2)\n"
+		"	la	%3,256(%3)\n"
+		"	la	%4,256(%4)\n"
+		"	la	%5,256(%5)\n"
+		"	brctg	0,0b\n"
+		"1:	ex	%0,0(1)\n"
+		"	ex	%0,6(1)\n"
+		"	ex	%0,12(1)\n"
+		"	ex	%0,18(1)\n"
+		"	j	3f\n"
+		"2:	xc	0(1,%1),0(%2)\n"
+		"	xc	0(1,%1),0(%3)\n"
+		"	xc	0(1,%1),0(%4)\n"
+		"	xc	0(1,%1),0(%5)\n"
+		"3:\n"
+		: "+d" (bytes), "+a" (p1), "+a" (p2), "+a" (p3), "+a" (p4),
+		  "+a" (reg7)
+		: : "0", "1", "cc", "memory");
+}
+
+struct xor_block_template xor_block_xc = {
+	.name = "xc",
+	.do_2 = xor_xc_2,
+	.do_3 = xor_xc_3,
+	.do_4 = xor_xc_4,
+	.do_5 = xor_xc_5,
+};
+EXPORT_SYMBOL(xor_block_xc);

arch/s390/mm/Makefile

···
 # Makefile for the linux s390-specific parts of the memory manager.
 #
 
-obj-y		:= init.o fault.o extmem.o mmap.o vmem.o pgtable.o maccess.o
+obj-y		:= init.o fault.o extmem.o mmap.o vmem.o maccess.o
 obj-y		+= page-states.o gup.o extable.o pageattr.o mem_detect.o
+obj-y		+= pgtable.o pgalloc.o
 
 obj-$(CONFIG_CMM)		+= cmm.o
 obj-$(CONFIG_HUGETLB_PAGE)	+= hugetlbpage.o
 obj-$(CONFIG_S390_PTDUMP)	+= dump_pagetables.o
+obj-$(CONFIG_PGSTE)		+= gmap.o

arch/s390/mm/extmem.c

···
 		goto out_free;
 	}
 	if (diag_cc > 1) {
-		pr_warning("Querying a DCSS type failed with rc=%ld\n", vmrc);
+		pr_warn("Querying a DCSS type failed with rc=%ld\n", vmrc);
 		rc = dcss_diag_translate_rc (vmrc);
 		goto out_free;
 	}
···
 		goto out_resource;
 	}
 	if (diag_cc > 1) {
-		pr_warning("Loading DCSS %s failed with rc=%ld\n", name,
-			   end_addr);
+		pr_warn("Loading DCSS %s failed with rc=%ld\n", name, end_addr);
 		rc = dcss_diag_translate_rc(end_addr);
 		dcss_diag(&purgeseg_scode, seg->dcss_name,
 				&dummy, &dummy);
···
 		goto out_unlock;
 	}
 	if (atomic_read (&seg->ref_count) != 1) {
-		pr_warning("DCSS %s is in use and cannot be reloaded\n",
-			   name);
+		pr_warn("DCSS %s is in use and cannot be reloaded\n", name);
 		rc = -EAGAIN;
 		goto out_unlock;
 	}
···
 			seg->res->flags |= IORESOURCE_READONLY;
 
 	if (request_resource(&iomem_resource, seg->res)) {
-		pr_warning("DCSS %s overlaps with used memory resources "
-			   "and cannot be reloaded\n", name);
+		pr_warn("DCSS %s overlaps with used memory resources and cannot be reloaded\n",
+			name);
 		rc = -EBUSY;
 		kfree(seg->res);
 		goto out_del_mem;
···
 		goto out_del_res;
 	}
 	if (diag_cc > 1) {
-		pr_warning("Reloading DCSS %s failed with rc=%ld\n", name,
-			   end_addr);
+		pr_warn("Reloading DCSS %s failed with rc=%ld\n",
+			name, end_addr);
 		rc = dcss_diag_translate_rc(end_addr);
 		goto out_del_res;
 	}

arch/s390/mm/fault.c

···
 #include <asm/asm-offsets.h>
 #include <asm/diag.h>
 #include <asm/pgtable.h>
+#include <asm/gmap.h>
 #include <asm/irq.h>
 #include <asm/mmu_context.h>
 #include <asm/facility.h>
···
 {
 	unsigned long asce;
 
+	pr_alert("Failing address: %016lx TEID: %016lx\n",
+		 regs->int_parm_long & __FAIL_ADDR_MASK, regs->int_parm_long);
 	pr_alert("Fault in ");
 	switch (regs->int_parm_long & 3) {
 	case 3:
···
 	dump_pagetable(asce, regs->int_parm_long & __FAIL_ADDR_MASK);
 }
 
-static inline void report_user_fault(struct pt_regs *regs, long signr)
+int show_unhandled_signals = 1;
+
+void report_user_fault(struct pt_regs *regs, long signr, int is_mm_fault)
 {
 	if ((task_pid_nr(current) > 1) && !show_unhandled_signals)
 		return;
···
 	       regs->int_code & 0xffff, regs->int_code >> 17);
 	print_vma_addr(KERN_CONT "in ", regs->psw.addr);
 	printk(KERN_CONT "\n");
-	printk(KERN_ALERT "failing address: %016lx TEID: %016lx\n",
-	       regs->int_parm_long & __FAIL_ADDR_MASK, regs->int_parm_long);
-	dump_fault_info(regs);
+	if (is_mm_fault)
+		dump_fault_info(regs);
 	show_regs(regs);
 }
 
···
 {
 	struct siginfo si;
 
-	report_user_fault(regs, SIGSEGV);
+	report_user_fault(regs, SIGSEGV, 1);
 	si.si_signo = SIGSEGV;
 	si.si_code = si_code;
 	si.si_addr = (void __user *)(regs->int_parm_long & __FAIL_ADDR_MASK);
···
 	else
 		printk(KERN_ALERT "Unable to handle kernel paging request"
 		       " in virtual user address space\n");
-	printk(KERN_ALERT "failing address: %016lx TEID: %016lx\n",
-	       regs->int_parm_long & __FAIL_ADDR_MASK, regs->int_parm_long);
 	dump_fault_info(regs);
 	die(regs, "Oops");
 	do_exit(SIGKILL);

arch/s390/mm/gmap.c

···
+/*
+ *  KVM guest address space mapping code
+ *
+ *    Copyright IBM Corp. 2007, 2016
+ *    Author(s): Martin Schwidefsky <[e-mail]>
+ */
+
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/swap.h>
+#include <linux/smp.h>
+#include <linux/spinlock.h>
+#include <linux/slab.h>
+#include <linux/swapops.h>
+#include <linux/ksm.h>
+#include <linux/mman.h>
+
+#include <asm/pgtable.h>
+#include <asm/pgalloc.h>
+#include <asm/gmap.h>
+#include <asm/tlb.h>
+
+/**
+ * gmap_alloc - allocate a guest address space
+ * @mm: pointer to the parent mm_struct
+ * @limit: maximum size of the gmap address space
+ *
+ * Returns a guest address space structure.
+ */
+struct gmap *gmap_alloc(struct mm_struct *mm, unsigned long limit)
+{
+	struct gmap *gmap;
+	struct page *page;
+	unsigned long *table;
+	unsigned long etype, atype;
+
+	if (limit < (1UL << 31)) {
+		limit = (1UL << 31) - 1;
+		atype = _ASCE_TYPE_SEGMENT;
+		etype = _SEGMENT_ENTRY_EMPTY;
+	} else if (limit < (1UL << 42)) {
+		limit = (1UL << 42) - 1;
+		atype = _ASCE_TYPE_REGION3;
+		etype = _REGION3_ENTRY_EMPTY;
+	} else if (limit < (1UL << 53)) {
+		limit = (1UL << 53) - 1;
+		atype = _ASCE_TYPE_REGION2;
+		etype = _REGION2_ENTRY_EMPTY;
+	} else {
+		limit = -1UL;
+		atype = _ASCE_TYPE_REGION1;
+		etype = _REGION1_ENTRY_EMPTY;
+	}
+	gmap = kzalloc(sizeof(struct gmap), GFP_KERNEL);
+	if (!gmap)
+		goto out;
+	INIT_LIST_HEAD(&gmap->crst_list);
+	INIT_RADIX_TREE(&gmap->guest_to_host, GFP_KERNEL);
+	INIT_RADIX_TREE(&gmap->host_to_guest, GFP_ATOMIC);
+	spin_lock_init(&gmap->guest_table_lock);
+	gmap->mm = mm;
+	page = alloc_pages(GFP_KERNEL, 2);
+	if (!page)
+		goto out_free;
+	page->index = 0;
+	list_add(&page->lru, &gmap->crst_list);
+	table = (unsigned long *) page_to_phys(page);
+	crst_table_init(table, etype);
+	gmap->table = table;
+	gmap->asce = atype | _ASCE_TABLE_LENGTH |
+		_ASCE_USER_BITS | __pa(table);
+	gmap->asce_end = limit;
+	down_write(&mm->mmap_sem);
+	list_add(&gmap->list, &mm->context.gmap_list);
+	up_write(&mm->mmap_sem);
+	return gmap;
+
+out_free:
+	kfree(gmap);
+out:
+	return NULL;
+}
+EXPORT_SYMBOL_GPL(gmap_alloc);
+
+static void gmap_flush_tlb(struct gmap *gmap)
+{
+	if (MACHINE_HAS_IDTE)
+		__tlb_flush_asce(gmap->mm, gmap->asce);
+	else
+		__tlb_flush_global();
+}
+
+static void gmap_radix_tree_free(struct radix_tree_root *root)
+{
+	struct radix_tree_iter iter;
+	unsigned long indices[16];
+	unsigned long index;
+	void **slot;
+	int i, nr;
+
+	/* A radix tree is freed by deleting all of its entries */
+	index = 0;
+	do {
+		nr = 0;
+		radix_tree_for_each_slot(slot, root, &iter, index) {
+			indices[nr] = iter.index;
+			if (++nr == 16)
+				break;
+		}
+		for (i = 0; i < nr; i++) {
+			index = indices[i];
+			radix_tree_delete(root, index);
+		}
+	} while (nr > 0);
+}
+
+/**
+ * gmap_free - free a guest address space
+ * @gmap: pointer to the guest address space structure
+ */
+void gmap_free(struct gmap *gmap)
+{
+	struct page *page, *next;
+
+	/* Flush tlb. */
+	if (MACHINE_HAS_IDTE)
+		__tlb_flush_asce(gmap->mm, gmap->asce);
+	else
+		__tlb_flush_global();
+
+	/* Free all segment & region tables. */
+	list_for_each_entry_safe(page, next, &gmap->crst_list, lru)
+		__free_pages(page, 2);
+	gmap_radix_tree_free(&gmap->guest_to_host);
+	gmap_radix_tree_free(&gmap->host_to_guest);
+	down_write(&gmap->mm->mmap_sem);
+	list_del(&gmap->list);
+	up_write(&gmap->mm->mmap_sem);
+	kfree(gmap);
+}
+EXPORT_SYMBOL_GPL(gmap_free);
+
+/**
+ * gmap_enable - switch primary space to the guest address space
+ * @gmap: pointer to the guest address space structure
+ */
+void gmap_enable(struct gmap *gmap)
+{
+	S390_lowcore.gmap = (unsigned long) gmap;
+}
+EXPORT_SYMBOL_GPL(gmap_enable);
+
+/**
+ * gmap_disable - switch back to the standard primary address space
+ * @gmap: pointer to the guest address space structure
+ */
+void gmap_disable(struct gmap *gmap)
+{
+	S390_lowcore.gmap = 0UL;
+}
+EXPORT_SYMBOL_GPL(gmap_disable);
+
+/*
+ * gmap_alloc_table is assumed to be called with mmap_sem held
+ */
+static int gmap_alloc_table(struct gmap *gmap, unsigned long *table,
+			    unsigned long init, unsigned long gaddr)
+{
+	struct page *page;
+	unsigned long *new;
+
+	/* since we dont free the gmap table until gmap_free we can unlock */
+	page = alloc_pages(GFP_KERNEL, 2);
+	if (!page)
+		return -ENOMEM;
+	new = (unsigned long *) page_to_phys(page);
+	crst_table_init(new, init);
+	spin_lock(&gmap->mm->page_table_lock);
+	if (*table & _REGION_ENTRY_INVALID) {
+		list_add(&page->lru, &gmap->crst_list);
+		*table = (unsigned long) new | _REGION_ENTRY_LENGTH |
+			(*table & _REGION_ENTRY_TYPE_MASK);
+		page->index = gaddr;
+		page = NULL;
+	}
+	spin_unlock(&gmap->mm->page_table_lock);
+	if (page)
+		__free_pages(page, 2);
+	return 0;
+}
+
+/**
+ * __gmap_segment_gaddr - find virtual address from segment pointer
+ * @entry: pointer to a segment table entry in the guest address space
+ *
+ * Returns the virtual address in the guest address space for the segment
+ */
+static unsigned long __gmap_segment_gaddr(unsigned long *entry)
+{
+	struct page *page;
+	unsigned long offset, mask;
+
+	offset = (unsigned long) entry / sizeof(unsigned long);
+	offset = (offset & (PTRS_PER_PMD - 1)) * PMD_SIZE;
+	mask = ~(PTRS_PER_PMD * sizeof(pmd_t) - 1);
+	page = virt_to_page((void *)((unsigned long) entry & mask));
+	return page->index + offset;
+}
+
+/**
+ * __gmap_unlink_by_vmaddr - unlink a single segment via a host address
+ * @gmap: pointer to the guest address space structure
+ * @vmaddr: address in the host process address space
+ *
+ * Returns 1 if a TLB flush is required
+ */
+static int __gmap_unlink_by_vmaddr(struct gmap *gmap, unsigned long vmaddr)
+{
+	unsigned long *entry;
+	int flush = 0;
+
+	spin_lock(&gmap->guest_table_lock);
+	entry = radix_tree_delete(&gmap->host_to_guest, vmaddr >> PMD_SHIFT);
+	if (entry) {
+		flush = (*entry != _SEGMENT_ENTRY_INVALID);
+		*entry = _SEGMENT_ENTRY_INVALID;
+	}
+	spin_unlock(&gmap->guest_table_lock);
+	return flush;
+}
+
+/**
+ * __gmap_unmap_by_gaddr - unmap a single segment via a guest address
+ * @gmap: pointer to the guest address space structure
+ * @gaddr: address in the guest address space
+ *
+ * Returns 1 if a TLB flush is required
+ */
+static int __gmap_unmap_by_gaddr(struct gmap *gmap, unsigned long gaddr)
+{
+	unsigned long vmaddr;
+
+	vmaddr = (unsigned long) radix_tree_delete(&gmap->guest_to_host,
+						   gaddr >> PMD_SHIFT);
+	return vmaddr ? __gmap_unlink_by_vmaddr(gmap, vmaddr) : 0;
+}
+
+/**
+ * gmap_unmap_segment - unmap segment from the guest address space
+ * @gmap: pointer to the guest address space structure
+ * @to: address in the guest address space
+ * @len: length of the memory area to unmap
+ *
+ * Returns 0 if the unmap succeeded, -EINVAL if not.
+ */
+int gmap_unmap_segment(struct gmap *gmap, unsigned long to, unsigned long len)
+{
+	unsigned long off;
+	int flush;
+
+	if ((to | len) & (PMD_SIZE - 1))
+		return -EINVAL;
+	if (len == 0 || to + len < to)
+		return -EINVAL;
+
+	flush = 0;
+	down_write(&gmap->mm->mmap_sem);
+	for (off = 0; off < len; off += PMD_SIZE)
+		flush |= __gmap_unmap_by_gaddr(gmap, to + off);
+	up_write(&gmap->mm->mmap_sem);
+	if (flush)
+		gmap_flush_tlb(gmap);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(gmap_unmap_segment);
+
+/**
+ * gmap_map_segment - map a segment to the guest address space
+ * @gmap: pointer to the guest address space structure
+ * @from: source address in the parent address space
+ * @to: target address in the guest address space
+ * @len: length of the memory area to map
+ *
+ * Returns 0 if the mmap succeeded, -EINVAL or -ENOMEM if not.
+ */
+int gmap_map_segment(struct gmap *gmap, unsigned long from,
+		     unsigned long to, unsigned long len)
+{
+	unsigned long off;
+	int flush;
+
+	if ((from | to | len) & (PMD_SIZE - 1))
+		return -EINVAL;
+	if (len == 0 || from + len < from || to + len < to ||
+	    from + len > TASK_MAX_SIZE || to + len > gmap->asce_end)
+		return -EINVAL;
+
+	flush = 0;
+	down_write(&gmap->mm->mmap_sem);
+	for (off = 0; off < len; off += PMD_SIZE) {
+		/* Remove old translation */
+		flush |= __gmap_unmap_by_gaddr(gmap, to + off);
+		/* Store new translation */
+		if (radix_tree_insert(&gmap->guest_to_host,
+				      (to + off) >> PMD_SHIFT,
+				      (void *) from + off))
+			break;
+	}
+	up_write(&gmap->mm->mmap_sem);
+	if (flush)
+		gmap_flush_tlb(gmap);
+	if (off >= len)
+		return 0;
+	gmap_unmap_segment(gmap, to, len);
+	return -ENOMEM;
+}
+EXPORT_SYMBOL_GPL(gmap_map_segment);
+
+/**
+ * __gmap_translate - translate a guest address to a user space address
+ * @gmap: pointer to guest mapping meta data structure
+ * @gaddr: guest address
+ *
+ * Returns user space address which corresponds to the guest address or
+ * -EFAULT if no such mapping exists.
+ * This function does not establish potentially missing page table entries.
+ * The mmap_sem of the mm that belongs to the address space must be held
+ * when this function gets called.
+ */
+unsigned long __gmap_translate(struct gmap *gmap, unsigned long gaddr)
+{
+	unsigned long vmaddr;
+
+	vmaddr = (unsigned long)
+		radix_tree_lookup(&gmap->guest_to_host, gaddr >> PMD_SHIFT);
+	return vmaddr ? (vmaddr | (gaddr & ~PMD_MASK)) : -EFAULT;
+}
+EXPORT_SYMBOL_GPL(__gmap_translate);
+
+/**
+ * gmap_translate - translate a guest address to a user space address
+ * @gmap: pointer to guest mapping meta data structure
+ * @gaddr: guest address
+ *
+ * Returns user space address which corresponds to the guest address or
+ * -EFAULT if no such mapping exists.
+ * This function does not establish potentially missing page table entries.
+ */
+unsigned long gmap_translate(struct gmap *gmap, unsigned long gaddr)
+{
+	unsigned long rc;
+
+	down_read(&gmap->mm->mmap_sem);
+	rc = __gmap_translate(gmap, gaddr);
+	up_read(&gmap->mm->mmap_sem);
+	return rc;
+}
+EXPORT_SYMBOL_GPL(gmap_translate);
+
+/**
+ * gmap_unlink - disconnect a page table from the gmap shadow tables
+ * @gmap: pointer to guest mapping meta data structure
+ * @table: pointer to the host page table
+ * @vmaddr: vm address associated with the host page table
+ */
+void gmap_unlink(struct mm_struct *mm, unsigned long *table,
+		 unsigned long vmaddr)
+{
+	struct gmap *gmap;
+	int flush;
+
+	list_for_each_entry(gmap, &mm->context.gmap_list, list) {
+		flush = __gmap_unlink_by_vmaddr(gmap, vmaddr);
+		if (flush)
+			gmap_flush_tlb(gmap);
+	}
+}
+
+/**
+ * gmap_link - set up shadow page tables to connect a host to a guest address
+ * @gmap: pointer to guest mapping meta data structure
+ * @gaddr: guest address
+ * @vmaddr: vm address
+ *
+ * Returns 0 on success, -ENOMEM for out of memory conditions, and -EFAULT
+ * if the vm address is already mapped to a different guest segment.
+ * The mmap_sem of the mm that belongs to the address space must be held
+ * when this function gets called.
+ */
+int __gmap_link(struct gmap *gmap, unsigned long gaddr, unsigned long vmaddr)
+{
+	struct mm_struct *mm;
+	unsigned long *table;
+	spinlock_t *ptl;
+	pgd_t *pgd;
+	pud_t *pud;
+	pmd_t *pmd;
+	int rc;
+
+	/* Create higher level tables in the gmap page table */
+	table = gmap->table;
+	if ((gmap->asce & _ASCE_TYPE_MASK) >= _ASCE_TYPE_REGION1) {
+		table += (gaddr >> 53) & 0x7ff;
+		if ((*table & _REGION_ENTRY_INVALID) &&
+		    gmap_alloc_table(gmap, table, _REGION2_ENTRY_EMPTY,
+				     gaddr & 0xffe0000000000000UL))
+			return -ENOMEM;
+		table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+	}
+	if ((gmap->asce & _ASCE_TYPE_MASK) >= _ASCE_TYPE_REGION2) {
+		table += (gaddr >> 42) & 0x7ff;
+		if ((*table & _REGION_ENTRY_INVALID) &&
+		    gmap_alloc_table(gmap, table, _REGION3_ENTRY_EMPTY,
+				     gaddr & 0xfffffc0000000000UL))
+			return -ENOMEM;
+		table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+	}
+	if ((gmap->asce & _ASCE_TYPE_MASK) >= _ASCE_TYPE_REGION3) {
+		table += (gaddr >> 31) & 0x7ff;
+		if ((*table & _REGION_ENTRY_INVALID) &&
+		    gmap_alloc_table(gmap, table, _SEGMENT_ENTRY_EMPTY,
+				     gaddr & 0xffffffff80000000UL))
+			return -ENOMEM;
+		table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+	}
+	table += (gaddr >> 20) & 0x7ff;
+	/* Walk the parent mm page table */
+	mm = gmap->mm;
+	pgd = pgd_offset(mm, vmaddr);
+	VM_BUG_ON(pgd_none(*pgd));
+	pud = pud_offset(pgd, vmaddr);
+	VM_BUG_ON(pud_none(*pud));
+	pmd = pmd_offset(pud, vmaddr);
+	VM_BUG_ON(pmd_none(*pmd));
+	/* large pmds cannot yet be handled */
+	if (pmd_large(*pmd))
+		return -EFAULT;
+	/* Link gmap segment table entry location to page table. */
+	rc = radix_tree_preload(GFP_KERNEL);
+	if (rc)
+		return rc;
+	ptl = pmd_lock(mm, pmd);
+	spin_lock(&gmap->guest_table_lock);
+	if (*table == _SEGMENT_ENTRY_INVALID) {
+		rc = radix_tree_insert(&gmap->host_to_guest,
+				       vmaddr >> PMD_SHIFT, table);
+		if (!rc)
+			*table = pmd_val(*pmd);
+	} else
+		rc = 0;
+	spin_unlock(&gmap->guest_table_lock);
+	spin_unlock(ptl);
+	radix_tree_preload_end();
+	return rc;
+}
+
+/**
+ * gmap_fault - resolve a fault on a guest address
+ * @gmap: pointer to guest mapping meta data structure
+ * @gaddr: guest address
+ * @fault_flags: flags to pass down to handle_mm_fault()
+ *
+ * Returns 0 on success, -ENOMEM for out of memory conditions, and -EFAULT
+ * if the vm address is already mapped to a different guest segment.
+ */
+int gmap_fault(struct gmap *gmap, unsigned long gaddr,
+	       unsigned int fault_flags)
+{
+	unsigned long vmaddr;
+	int rc;
+	bool unlocked;
+
+	down_read(&gmap->mm->mmap_sem);
+
+retry:
+	unlocked = false;
+	vmaddr = __gmap_translate(gmap, gaddr);
+	if (IS_ERR_VALUE(vmaddr)) {
+		rc = vmaddr;
+		goto out_up;
+	}
+	if (fixup_user_fault(current, gmap->mm, vmaddr, fault_flags,
+			     &unlocked)) {
+		rc = -EFAULT;
+		goto out_up;
+	}
+	/*
+	 * In the case that fixup_user_fault unlocked the mmap_sem during
+	 * faultin redo __gmap_translate to not race with a map/unmap_segment.
+	 */
+	if (unlocked)
+		goto retry;
+
+	rc = __gmap_link(gmap, gaddr, vmaddr);
+out_up:
+	up_read(&gmap->mm->mmap_sem);
+	return rc;
+}
+EXPORT_SYMBOL_GPL(gmap_fault);
+
+/*
+ * this function is assumed to be called with mmap_sem held
+ */
+void __gmap_zap(struct gmap *gmap, unsigned long gaddr)
+{
+	unsigned long vmaddr;
+	spinlock_t *ptl;
+	pte_t *ptep;
+
+	/* Find the vm address for the guest address */
+	vmaddr = (unsigned long) radix_tree_lookup(&gmap->guest_to_host,
+						   gaddr >> PMD_SHIFT);
+	if (vmaddr) {
+		vmaddr |= gaddr & ~PMD_MASK;
+		/* Get pointer to the page table entry */
+		ptep = get_locked_pte(gmap->mm, vmaddr, &ptl);
+		if (likely(ptep))
+			ptep_zap_unused(gmap->mm, vmaddr, ptep, 0);
+		pte_unmap_unlock(ptep, ptl);
+	}
+}
+EXPORT_SYMBOL_GPL(__gmap_zap);
+
+void gmap_discard(struct gmap *gmap, unsigned long from, unsigned long to)
+{
+	unsigned long gaddr, vmaddr, size;
+	struct vm_area_struct *vma;
+
+	down_read(&gmap->mm->mmap_sem);
+	for (gaddr = from; gaddr < to;
+	     gaddr = (gaddr + PMD_SIZE) & PMD_MASK) {
+		/* Find the vm address for the guest address */
+		vmaddr = (unsigned long)
+			radix_tree_lookup(&gmap->guest_to_host,
+					  gaddr >> PMD_SHIFT);
+		if (!vmaddr)
+			continue;
+		vmaddr |= gaddr & ~PMD_MASK;
+		/* Find vma in the parent mm */
+		vma = find_vma(gmap->mm, vmaddr);
+		size = min(to - gaddr, PMD_SIZE - (gaddr & ~PMD_MASK));
+		zap_page_range(vma, vmaddr, size, NULL);
+	}
+	up_read(&gmap->mm->mmap_sem);
+}
+EXPORT_SYMBOL_GPL(gmap_discard);
+
+static LIST_HEAD(gmap_notifier_list);
+static DEFINE_SPINLOCK(gmap_notifier_lock);
+
+/**
+ * gmap_register_ipte_notifier - register a pte invalidation callback
+ * @nb: pointer to the gmap notifier block
+ */
+void gmap_register_ipte_notifier(struct gmap_notifier *nb)
+{
+	spin_lock(&gmap_notifier_lock);
+	list_add(&nb->list, &gmap_notifier_list);
+	spin_unlock(&gmap_notifier_lock);
+}
+EXPORT_SYMBOL_GPL(gmap_register_ipte_notifier);
+
+/**
+ * gmap_unregister_ipte_notifier - remove a pte invalidation callback
+ * @nb: pointer to the gmap notifier block
+ */
+void gmap_unregister_ipte_notifier(struct gmap_notifier *nb)
+{
+	spin_lock(&gmap_notifier_lock);
+	list_del_init(&nb->list);
+	spin_unlock(&gmap_notifier_lock);
+}
+EXPORT_SYMBOL_GPL(gmap_unregister_ipte_notifier);
+
+/**
+ * gmap_ipte_notify - mark a range of ptes for invalidation notification
+ * @gmap: pointer to guest mapping meta data structure
+ * @gaddr: virtual address in the guest address space
+ * @len: size of area
+ *
+ * Returns 0 if for each page in the given range a gmap mapping exists and
+ * the invalidation notification could be set. If the gmap mapping is missing
+ * for one or more pages -EFAULT is returned. If no memory could be allocated
+ * -ENOMEM is returned. This function establishes missing page table entries.
+ */
+int gmap_ipte_notify(struct gmap *gmap, unsigned long gaddr, unsigned long len)
+{
+	unsigned long addr;
+	spinlock_t *ptl;
+	pte_t *ptep;
+	bool unlocked;
+	int rc = 0;
+
+	if ((gaddr & ~PAGE_MASK) || (len & ~PAGE_MASK))
+		return -EINVAL;
+	down_read(&gmap->mm->mmap_sem);
+	while (len) {
+		unlocked = false;
+		/* Convert gmap address and connect the page tables */
+		addr = __gmap_translate(gmap, gaddr);
+		if (IS_ERR_VALUE(addr)) {
+			rc = addr;
+			break;
+		}
+		/* Get the page mapped */
+		if (fixup_user_fault(current, gmap->mm, addr, FAULT_FLAG_WRITE,
+				     &unlocked)) {
+			rc = -EFAULT;
+			break;
+		}
+		/* While trying to map mmap_sem got unlocked. Let us retry */
+		if (unlocked)
+			continue;
+		rc = __gmap_link(gmap, gaddr, addr);
+		if (rc)
+			break;
+		/* Walk the process page table, lock and get pte pointer */
+		ptep = get_locked_pte(gmap->mm, addr, &ptl);
+		VM_BUG_ON(!ptep);
+		/* Set notification bit in the pgste of the pte */
+		if ((pte_val(*ptep) & (_PAGE_INVALID | _PAGE_PROTECT)) == 0) {
+			ptep_set_notify(gmap->mm, addr, ptep);
+			gaddr += PAGE_SIZE;
+			len -= PAGE_SIZE;
+		}
+		pte_unmap_unlock(ptep, ptl);
+	}
+	up_read(&gmap->mm->mmap_sem);
+	return rc;
+}
+EXPORT_SYMBOL_GPL(gmap_ipte_notify);
+
+/**
+ * ptep_notify - call all invalidation callbacks for a specific pte.
+ * @mm: pointer to the process mm_struct
+ * @addr: virtual address in the process address space
+ * @pte: pointer to the page table entry
+ *
+ * This function is assumed to be called with the page table lock held
+ * for the pte to notify.
+ */
+void ptep_notify(struct mm_struct *mm, unsigned long vmaddr, pte_t *pte)
+{
+	unsigned long offset, gaddr;
+	unsigned long *table;
+	struct gmap_notifier *nb;
+	struct gmap *gmap;
+
+	offset = ((unsigned long) pte) & (255 * sizeof(pte_t));
+	offset = offset * (4096 / sizeof(pte_t));
+	spin_lock(&gmap_notifier_lock);
+	list_for_each_entry(gmap, &mm->context.gmap_list, list) {
+		table = radix_tree_lookup(&gmap->host_to_guest,
+					  vmaddr >> PMD_SHIFT);
+		if (!table)
+			continue;
+		gaddr = __gmap_segment_gaddr(table) + offset;
+		list_for_each_entry(nb, &gmap_notifier_list, list)
+			nb->notifier_call(gmap, gaddr);
+	}
+	spin_unlock(&gmap_notifier_lock);
+}
+EXPORT_SYMBOL_GPL(ptep_notify);
+
+static inline void thp_split_mm(struct mm_struct *mm)
+{
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+	struct vm_area_struct *vma;
+	unsigned long addr;
+
+	for (vma = mm->mmap; vma != NULL; vma = vma->vm_next) {
+		for (addr = vma->vm_start;
+		     addr < vma->vm_end;
+		     addr += PAGE_SIZE)
+			follow_page(vma, addr, FOLL_SPLIT);
+		vma->vm_flags &= ~VM_HUGEPAGE;
+		vma->vm_flags |= VM_NOHUGEPAGE;
+	}
+	mm->def_flags |= VM_NOHUGEPAGE;
+#endif
+}
+
+/*
+ * switch on pgstes for its userspace process (for kvm)
+ */
+int s390_enable_sie(void)
+{
+	struct mm_struct *mm = current->mm;
+
+	/* Do we have pgstes? if yes, we are done */
+	if (mm_has_pgste(mm))
+		return 0;
+	/* Fail if the page tables are 2K */
+	if (!mm_alloc_pgste(mm))
+		return -EINVAL;
+	down_write(&mm->mmap_sem);
+	mm->context.has_pgste = 1;
+	/* split thp mappings and disable thp for future mappings */
+	thp_split_mm(mm);
+	up_write(&mm->mmap_sem);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(s390_enable_sie);
+
+/*
+ * Enable storage key handling from now on and initialize the storage
+ * keys with the default key.
+ */
+static int __s390_enable_skey(pte_t *pte, unsigned long addr,
+			      unsigned long next, struct mm_walk *walk)
+{
+	/*
+	 * Remove all zero page mappings,
+	 * after establishing a policy to forbid zero page mappings
+	 * following faults for that page will get fresh anonymous pages
+	 */
+	if (is_zero_pfn(pte_pfn(*pte)))
+		ptep_xchg_direct(walk->mm, addr, pte, __pte(_PAGE_INVALID));
+	/* Clear storage key */
+	ptep_zap_key(walk->mm, addr, pte);
+	return 0;
+}
+
+int s390_enable_skey(void)
+{
+	struct mm_walk walk = { .pte_entry = __s390_enable_skey };
+	struct mm_struct *mm = current->mm;
+	struct vm_area_struct *vma;
+	int rc = 0;
+
+	down_write(&mm->mmap_sem);
+	if (mm_use_skey(mm))
+		goto out_up;
+
+	mm->context.use_skey = 1;
+	for (vma = mm->mmap; vma; vma = vma->vm_next) {
+		if (ksm_madvise(vma, vma->vm_start, vma->vm_end,
+				MADV_UNMERGEABLE, &vma->vm_flags)) {
+			mm->context.use_skey = 0;
+			rc = -ENOMEM;
+			goto out_up;
+		}
+	}
+	mm->def_flags &= ~VM_MERGEABLE;
+
+	walk.mm = mm;
+	walk_page_range(0, TASK_SIZE, &walk);
+
+out_up:
+	up_write(&mm->mmap_sem);
+	return rc;
+}
+EXPORT_SYMBOL_GPL(s390_enable_skey);
+
+/*
+ * Reset CMMA state, make all pages stable again.
+ */
+static int __s390_reset_cmma(pte_t *pte, unsigned long addr,
+			     unsigned long next, struct mm_walk *walk)
+{
+	ptep_zap_unused(walk->mm, addr, pte, 1);
+	return 0;
+}
+
+void s390_reset_cmma(struct mm_struct *mm)
+{
+	struct mm_walk walk = { .pte_entry = __s390_reset_cmma };
+
+	down_write(&mm->mmap_sem);
+	walk.mm = mm;
+	walk_page_range(0, TASK_SIZE, &walk);
+	up_write(&mm->mmap_sem);
+}
+EXPORT_SYMBOL_GPL(s390_reset_cmma);

arch/s390/mm/hugetlbpage.c

···
 			      unsigned long addr, pte_t *ptep)
 {
 	pmd_t *pmdp = (pmd_t *) ptep;
-	pte_t pte = huge_ptep_get(ptep);
+	pmd_t old;
 
-	pmdp_flush_direct(mm, addr, pmdp);
-	pmd_val(*pmdp) = _SEGMENT_ENTRY_EMPTY;
-	return pte;
+	old = pmdp_xchg_direct(mm, addr, pmdp, __pmd(_SEGMENT_ENTRY_EMPTY));
+	return __pmd_to_pte(old);
 }
 
 pte_t *huge_pte_alloc(struct mm_struct *mm,

arch/s390/mm/pageattr.c

···
 static void change_page_attr(unsigned long addr, int numpages,
 			     pte_t (*set) (pte_t))
 {
-	pte_t *ptep, pte;
+	pte_t *ptep;
 	int i;
 
 	for (i = 0; i < numpages; i++) {
 		ptep = walk_page_table(addr);
 		if (WARN_ON_ONCE(!ptep))
 			break;
-		pte = *ptep;
-		pte = set(pte);
-		__ptep_ipte(addr, ptep);
-		*ptep = pte;
+		*ptep = set(*ptep);
 		addr += PAGE_SIZE;
 	}
+	__tlb_flush_kernel();
 }
 
 int set_memory_ro(unsigned long addr, int numpages)

arch/s390/mm/pgalloc.c

···
+/*
+ *  Page table allocation functions
+ *
+ *    Copyright IBM Corp. 2016
+ *    Author(s): Martin Schwidefsky <[e-mail]>
+ */
+
+#include <linux/mm.h>
+#include <linux/sysctl.h>
+#include <asm/mmu_context.h>
+#include <asm/pgalloc.h>
+#include <asm/gmap.h>
+#include <asm/tlb.h>
+#include <asm/tlbflush.h>
+
+#ifdef CONFIG_PGSTE
+
+static int page_table_allocate_pgste_min = 0;
+static int page_table_allocate_pgste_max = 1;
+int page_table_allocate_pgste = 0;
+EXPORT_SYMBOL(page_table_allocate_pgste);
+
+static struct ctl_table page_table_sysctl[] = {
+	{
+		.procname	= "allocate_pgste",
+		.data		= &page_table_allocate_pgste,
+		.maxlen		= sizeof(int),
+		.mode		= S_IRUGO | S_IWUSR,
+		.proc_handler	= proc_dointvec,
+		.extra1		= &page_table_allocate_pgste_min,
+		.extra2		= &page_table_allocate_pgste_max,
+	},
+	{ }
+};
+
+static struct ctl_table page_table_sysctl_dir[] = {
+	{
+		.procname	= "vm",
+		.maxlen		= 0,
+		.mode		= 0555,
+		.child		= page_table_sysctl,
+	},
+	{ }
+};
+
+static int __init page_table_register_sysctl(void)
+{
+	return register_sysctl_table(page_table_sysctl_dir) ? 0 : -ENOMEM;
+}
+__initcall(page_table_register_sysctl);
+
+#endif /* CONFIG_PGSTE */
+
+unsigned long *crst_table_alloc(struct mm_struct *mm)
+{
+	struct page *page = alloc_pages(GFP_KERNEL, 2);
+
+	if (!page)
+		return NULL;
+	return (unsigned long *) page_to_phys(page);
+}
+
+void crst_table_free(struct mm_struct *mm, unsigned long *table)
+{
+	free_pages((unsigned long) table, 2);
+}
+
+static void __crst_table_upgrade(void *arg)
+{
+	struct mm_struct *mm = arg;
+
+	if (current->active_mm == mm) {
+		clear_user_asce();
+		set_user_asce(mm);
+	}
+	__tlb_flush_local();
+}
+
+int crst_table_upgrade(struct mm_struct *mm, unsigned long limit)
+{
+	unsigned long *table, *pgd;
+	unsigned long entry;
+	int flush;
+
+	BUG_ON(limit > TASK_MAX_SIZE);
+	flush = 0;
+repeat:
+	table = crst_table_alloc(mm);
+	if (!table)
+		return -ENOMEM;
+	spin_lock_bh(&mm->page_table_lock);
+	if (mm->context.asce_limit < limit) {
+		pgd = (unsigned long *) mm->pgd;
+		if (mm->context.asce_limit <= (1UL << 31)) {
+			entry = _REGION3_ENTRY_EMPTY;
+			mm->context.asce_limit = 1UL << 42;
+			mm->context.asce_bits = _ASCE_TABLE_LENGTH |
+						_ASCE_USER_BITS |
+						_ASCE_TYPE_REGION3;
+		} else {
+			entry = _REGION2_ENTRY_EMPTY;
+			mm->context.asce_limit = 1UL << 53;
+			mm->context.asce_bits = _ASCE_TABLE_LENGTH |
+						_ASCE_USER_BITS |
+						_ASCE_TYPE_REGION2;
+		}
+		crst_table_init(table, entry);
+		pgd_populate(mm, (pgd_t *) table, (pud_t *) pgd);
+		mm->pgd = (pgd_t *) table;
+		mm->task_size = mm->context.asce_limit;
+		table = NULL;
+		flush = 1;
+	}
+	spin_unlock_bh(&mm->page_table_lock);
+	if (table)
+		crst_table_free(mm, table);
+	if (mm->context.asce_limit < limit)
+		goto repeat;
+	if (flush)
+		on_each_cpu(__crst_table_upgrade, mm, 0);
+	return 0;
+}
+
+void crst_table_downgrade(struct mm_struct *mm, unsigned long limit)
+{
+	pgd_t *pgd;
+
+	if (current->active_mm == mm) {
+		clear_user_asce();
+		__tlb_flush_mm(mm);
+	}
+	while (mm->context.asce_limit > limit) {
+		pgd = mm->pgd;
+		switch (pgd_val(*pgd) & _REGION_ENTRY_TYPE_MASK) {
+		case _REGION_ENTRY_TYPE_R2:
+			mm->context.asce_limit = 1UL << 42;
+			mm->context.asce_bits = _ASCE_TABLE_LENGTH |
+						_ASCE_USER_BITS |
+						_ASCE_TYPE_REGION3;
+			break;
+		case _REGION_ENTRY_TYPE_R3:
+			mm->context.asce_limit = 1UL << 31;
+			mm->context.asce_bits = _ASCE_TABLE_LENGTH |
+						_ASCE_USER_BITS |
+						_ASCE_TYPE_SEGMENT;
+			break;
+		default:
+			BUG();
+		}
+		mm->pgd = (pgd_t *) (pgd_val(*pgd) & _REGION_ENTRY_ORIGIN);
+		mm->task_size = mm->context.asce_limit;
+		crst_table_free(mm, (unsigned long *) pgd);
+	}
+	if (current->active_mm == mm)
+		set_user_asce(mm);
+}
+
+static inline unsigned int atomic_xor_bits(atomic_t *v, unsigned int bits)
+{
+	unsigned int old, new;
+
+	do {
+		old = atomic_read(v);
+		new = old ^ bits;
+	} while (atomic_cmpxchg(v, old, new) != old);
+	return new;
+}
+
+/*
+ * page table entry allocation/free routines.
+ */
+unsigned long *page_table_alloc(struct mm_struct *mm)
+{
+	unsigned long *table;
+	struct page *page;
+	unsigned int mask, bit;
+
+	/* Try to get a fragment of a 4K page as a 2K page table */
+	if (!mm_alloc_pgste(mm)) {
+		table = NULL;
+		spin_lock_bh(&mm->context.list_lock);
+		if (!list_empty(&mm->context.pgtable_list)) {
+			page = list_first_entry(&mm->context.pgtable_list,
+						struct page, lru);
+			mask = atomic_read(&page->_mapcount);
+			mask = (mask | (mask >> 4)) & 3;
+			if (mask != 3) {
+				table = (unsigned long *) page_to_phys(page);
+				bit = mask & 1;		/* =1 -> second 2K */
+				if (bit)
+					table += PTRS_PER_PTE;
+				atomic_xor_bits(&page->_mapcount, 1U << bit);
+				list_del(&page->lru);
+			}
+		}
+		spin_unlock_bh(&mm->context.list_lock);
+		if (table)
+			return table;
+	}
+	/* Allocate a fresh page */
+	page = alloc_page(GFP_KERNEL|__GFP_REPEAT);
+	if (!page)
+		return NULL;
+	if (!pgtable_page_ctor(page)) {
+		__free_page(page);
+		return NULL;
+	}
+	/* Initialize page table */
+	table = (unsigned long *) page_to_phys(page);
+	if (mm_alloc_pgste(mm)) {
+		/* Return 4K page table with PGSTEs */
+		atomic_set(&page->_mapcount, 3);
+		clear_table(table, _PAGE_INVALID, PAGE_SIZE/2);
+		clear_table(table + PTRS_PER_PTE, 0, PAGE_SIZE/2);
+	} else {
+		/* Return the first 2K fragment of the page */
+		atomic_set(&page->_mapcount, 1);
+		clear_table(table, _PAGE_INVALID, PAGE_SIZE);
+		spin_lock_bh(&mm->context.list_lock);
+		list_add(&page->lru, &mm->context.pgtable_list);
+		spin_unlock_bh(&mm->context.list_lock);
+	}
+	return table;
+}
+
+void page_table_free(struct mm_struct *mm, unsigned long *table)
+{
+	struct page *page;
+	unsigned int bit, mask;
+
+	page = pfn_to_page(__pa(table) >> PAGE_SHIFT);
+	if (!mm_alloc_pgste(mm)) {
+		/* Free 2K page table fragment of a 4K page */
+		bit = (__pa(table) & ~PAGE_MASK)/(PTRS_PER_PTE*sizeof(pte_t));
+		spin_lock_bh(&mm->context.list_lock);
+		mask = atomic_xor_bits(&page->_mapcount, 1U << bit);
+		if (mask & 3)
+			list_add(&page->lru, &mm->context.pgtable_list);
+		else
+			list_del(&page->lru);
+		spin_unlock_bh(&mm->context.list_lock);
+		if (mask != 0)
+			return;
+	}
+
+	pgtable_page_dtor(page);
+	atomic_set(&page->_mapcount, -1);
+	__free_page(page);
+}
+
+void page_table_free_rcu(struct mmu_gather *tlb, unsigned long *table,
+			 unsigned long vmaddr)
+{
+	struct mm_struct *mm;
+	struct page *page;
+	unsigned int bit, mask;
+
+	mm = tlb->mm;
+	page = pfn_to_page(__pa(table) >> PAGE_SHIFT);
+	if (mm_alloc_pgste(mm)) {
+		gmap_unlink(mm, table, vmaddr);
+		table = (unsigned long *) (__pa(table) | 3);
+		tlb_remove_table(tlb, table);
+		return;
+	}
+	bit = (__pa(table) & ~PAGE_MASK) / (PTRS_PER_PTE*sizeof(pte_t));
+	spin_lock_bh(&mm->context.list_lock);
+	mask = atomic_xor_bits(&page->_mapcount, 0x11U << bit);
+	if (mask & 3)
+		list_add_tail(&page->lru, &mm->context.pgtable_list);
+	else
+		list_del(&page->lru);
+	spin_unlock_bh(&mm->context.list_lock);
+	table = (unsigned long *) (__pa(table) | (1U << bit));
+	tlb_remove_table(tlb, table);
+}
+
+static void __tlb_remove_table(void *_table)
+{
+	unsigned int mask = (unsigned long) _table & 3;
+	void *table = (void *)((unsigned long) _table ^ mask);
+	struct page *page = pfn_to_page(__pa(table) >> PAGE_SHIFT);
+
+	switch (mask) {
+	case 0:		/* pmd or pud */
+		free_pages((unsigned long) table, 2);
+		break;
+	case 1:		/* lower 2K of a 4K page table */
+	case 2:		/* higher 2K of a 4K page table */
+		if (atomic_xor_bits(&page->_mapcount, mask << 4) != 0)
+			break;
+		/* fallthrough */
+	case 3:		/* 4K page table with pgstes */
+		pgtable_page_dtor(page);
+		atomic_set(&page->_mapcount, -1);
+		__free_page(page);
+		break;
+	}
+}
+
+static void tlb_remove_table_smp_sync(void *arg)
+{
+	/* Simply deliver the interrupt */
+}
+
+static void tlb_remove_table_one(void *table)
+{
+	/*
+	 * This isn't an RCU grace period and hence the page-tables cannot be
+	 * assumed to be actually RCU-freed.
+	 *
+	 * It is however sufficient for software page-table walkers that rely
+	 * on IRQ disabling. See the comment near struct mmu_table_batch.
+	 */
+	smp_call_function(tlb_remove_table_smp_sync, NULL, 1);
+	__tlb_remove_table(table);
+}
+
+static void tlb_remove_table_rcu(struct rcu_head *head)
+{
+	struct mmu_table_batch *batch;
+	int i;
+
+	batch = container_of(head, struct mmu_table_batch, rcu);
+
+	for (i = 0; i < batch->nr; i++)
+		__tlb_remove_table(batch->tables[i]);
+
+	free_page((unsigned long)batch);
+}
+
+void tlb_table_flush(struct mmu_gather *tlb)
+{
+	struct mmu_table_batch **batch = &tlb->batch;
+
+	if (*batch) {
+		call_rcu_sched(&(*batch)->rcu, tlb_remove_table_rcu);
+		*batch = NULL;
+	}
+}
+
+void tlb_remove_table(struct mmu_gather *tlb, void *table)
+{
+	struct mmu_table_batch **batch = &tlb->batch;
+
+	tlb->mm->context.flush_mm = 1;
+	if (*batch == NULL) {
+		*batch = (struct mmu_table_batch *)
+			__get_free_page(GFP_NOWAIT | __GFP_NOWARN);
+		if (*batch == NULL) {
+			__tlb_flush_mm_lazy(tlb->mm);
+			tlb_remove_table_one(table);
+			return;
+		}
+		(*batch)->nr = 0;
+	}
+	(*batch)->tables[(*batch)->nr++] = table;
+	if ((*batch)->nr == MAX_TABLE_BATCH)
+		tlb_flush_mmu(tlb);
+}

arch/s390/mm/pgtable.c

···
 #include <asm/tlbflush.h>
 #include <asm/mmu_context.h>
 
-unsigned long *crst_table_alloc(struct mm_struct *mm)
+static inline pte_t ptep_flush_direct(struct mm_struct *mm,
+				      unsigned long addr, pte_t *ptep)
 {
-	struct page *page = alloc_pages(GFP_KERNEL, 2);
+	int active, count;
+	pte_t old;
 
-	if (!page)
-		return NULL;
-	return (unsigned long *) page_to_phys(page);
-}
-
-void crst_table_free(struct mm_struct *mm, unsigned long *table)
-{
-	free_pages((unsigned long) table, 2);
-}
-
-static void __crst_table_upgrade(void *arg)
-{
-	struct mm_struct *mm = arg;
-
-	if (current->active_mm == mm) {
-		clear_user_asce();
-		set_user_asce(mm);
-	}
-	__tlb_flush_local();
-}
-
-int crst_table_upgrade(struct mm_struct *mm, unsigned long limit)
-{
-	unsigned long *table, *pgd;
-	unsigned long entry;
-	int flush;
-
-	BUG_ON(limit > TASK_MAX_SIZE);
-	flush = 0;
-repeat:
-	table = crst_table_alloc(mm);
-	if (!table)
-		return -ENOMEM;
-	spin_lock_bh(&mm->page_table_lock);
-	if (mm->context.asce_limit < limit) {
-		pgd = (unsigned long *) mm->pgd;
-		if (mm->context.asce_limit <= (1UL << 31)) {
-			entry = _REGION3_ENTRY_EMPTY;
-			mm->context.asce_limit = 1UL << 42;
-			mm->context.asce_bits = _ASCE_TABLE_LENGTH |
-						_ASCE_USER_BITS |
-						_ASCE_TYPE_REGION3;
-		} else {
-			entry = _REGION2_ENTRY_EMPTY;
-			mm->context.asce_limit = 1UL << 53;
-			mm->context.asce_bits = _ASCE_TABLE_LENGTH |
-						_ASCE_USER_BITS |
-						_ASCE_TYPE_REGION2;
-		}
-		crst_table_init(table, entry);
-		pgd_populate(mm, (pgd_t *) table, (pud_t *) pgd);
-		mm->pgd = (pgd_t *) table;
-		mm->task_size = mm->context.asce_limit;
-		table = NULL;
-		flush = 1;
-	}
-	spin_unlock_bh(&mm->page_table_lock);
-	if (table)
-		crst_table_free(mm, table);
-	if (mm->context.asce_limit < limit)
-		goto repeat;
-	if (flush)
-		on_each_cpu(__crst_table_upgrade, mm, 0);
-	return 0;
-}
-
-void crst_table_downgrade(struct mm_struct *mm, unsigned long limit)
-{
-	pgd_t *pgd;
-
-	if (current->active_mm == mm) {
-		clear_user_asce();
-		__tlb_flush_mm(mm);
-	}
-	while (mm->context.asce_limit > limit) {
-		pgd = mm->pgd;
-		switch (pgd_val(*pgd) & _REGION_ENTRY_TYPE_MASK) {
-		case _REGION_ENTRY_TYPE_R2:
-			mm->context.asce_limit = 1UL << 42;
-			mm->context.asce_bits = _ASCE_TABLE_LENGTH |
-						_ASCE_USER_BITS |
-						_ASCE_TYPE_REGION3;
-			break;
-		case _REGION_ENTRY_TYPE_R3:
-			mm->context.asce_limit = 1UL << 31;
-			mm->context.asce_bits = _ASCE_TABLE_LENGTH |
-						_ASCE_USER_BITS |
-						_ASCE_TYPE_SEGMENT;
-			break;
-		default:
-			BUG();
-		}
-		mm->pgd = (pgd_t *) (pgd_val(*pgd) & _REGION_ENTRY_ORIGIN);
-		mm->task_size = mm->context.asce_limit;
-		crst_table_free(mm, (unsigned long *) pgd);
-	}
-	if (current->active_mm == mm)
-		set_user_asce(mm);
-}
-
-#ifdef CONFIG_PGSTE
-
-/**
- * gmap_alloc - allocate a guest address space
- * @mm: pointer to the parent mm_struct
- * @limit: maximum address of the gmap address space
- *
- * Returns a guest address space structure.
- */
-struct gmap *gmap_alloc(struct mm_struct *mm, unsigned long limit)
-{
-	struct gmap *gmap;
-	struct page *page;
-	unsigned long *table;
-	unsigned long etype, atype;
-
-	if (limit < (1UL << 31)) {
-		limit = (1UL << 31) - 1;
-		atype = _ASCE_TYPE_SEGMENT;
-		etype = _SEGMENT_ENTRY_EMPTY;
-	} else if (limit < (1UL << 42)) {
-		limit = (1UL << 42) - 1;
-		atype = _ASCE_TYPE_REGION3;
-		etype = _REGION3_ENTRY_EMPTY;
-	} else if (limit < (1UL << 53)) {
-		limit = (1UL << 53) - 1;
-		atype = _ASCE_TYPE_REGION2;
-		etype = _REGION2_ENTRY_EMPTY;
-	} else {
-		limit = -1UL;
-		atype = _ASCE_TYPE_REGION1;
-		etype = _REGION1_ENTRY_EMPTY;
-	}
-	gmap = kzalloc(sizeof(struct gmap), GFP_KERNEL);
-	if (!gmap)
-		goto out;
-	INIT_LIST_HEAD(&gmap->crst_list);
-	INIT_RADIX_TREE(&gmap->guest_to_host, GFP_KERNEL);
-	INIT_RADIX_TREE(&gmap->host_to_guest, GFP_ATOMIC);
-	spin_lock_init(&gmap->guest_table_lock);
-	gmap->mm = mm;
-	page = alloc_pages(GFP_KERNEL, 2);
-	if (!page)
-		goto out_free;
-	page->index = 0;
-	list_add(&page->lru, &gmap->crst_list);
-	table = (unsigned long *) page_to_phys(page);
-	crst_table_init(table, etype);
-	gmap->table = table;
-	gmap->asce = atype | _ASCE_TABLE_LENGTH |
-		_ASCE_USER_BITS | __pa(table);
-	gmap->asce_end = limit;
-	down_write(&mm->mmap_sem);
-	list_add(&gmap->list, &mm->context.gmap_list);
-	up_write(&mm->mmap_sem);
-	return gmap;
-
-out_free:
-	kfree(gmap);
-out:
-	return NULL;
-}
-EXPORT_SYMBOL_GPL(gmap_alloc);
-
-static void gmap_flush_tlb(struct gmap *gmap)
-{
-	if (MACHINE_HAS_IDTE)
-		__tlb_flush_asce(gmap->mm, gmap->asce);
+	old = *ptep;
+	if (unlikely(pte_val(old) & _PAGE_INVALID))
+		return old;
+	active = (mm == current->active_mm) ? 1 : 0;
+	count = atomic_add_return(0x10000, &mm->context.attach_count);
+	if (MACHINE_HAS_TLB_LC && (count & 0xffff) <= active &&
+	    cpumask_equal(mm_cpumask(mm), cpumask_of(smp_processor_id())))
+		__ptep_ipte_local(addr, ptep);
 	else
-		__tlb_flush_global();
+		__ptep_ipte(addr, ptep);
+	atomic_sub(0x10000, &mm->context.attach_count);
+	return old;
 }
 
-static void gmap_radix_tree_free(struct radix_tree_root *root)
+static inline pte_t ptep_flush_lazy(struct mm_struct *mm,
+				    unsigned long addr, pte_t *ptep)
 {
-	struct radix_tree_iter iter;
-	unsigned long indices[16];
-	unsigned long index;
-	void **slot;
-	int i, nr;
+	int active, count;
+	pte_t old;
 
-	/* A radix tree is freed by deleting all of its entries */
-	index = 0;
-	do {
-		nr = 0;
-		radix_tree_for_each_slot(slot, root, &iter, index) {
-			indices[nr] = iter.index;
-			if (++nr == 16)
-				break;
-		}
-		for (i = 0; i < nr; i++) {
-			index = indices[i];
-			radix_tree_delete(root, index);
-		}
-	} while (nr > 0);
-}
-
-/**
- * gmap_free - free a guest address space
- * @gmap: pointer to the guest address space structure
- */
-void gmap_free(struct gmap *gmap)
-{
-	struct page *page, *next;
-
-	/* Flush tlb. */
-	if (MACHINE_HAS_IDTE)
-		__tlb_flush_asce(gmap->mm, gmap->asce);
-	else
-		__tlb_flush_global();
-
-	/* Free all segment & region tables. */
-	list_for_each_entry_safe(page, next, &gmap->crst_list, lru)
-		__free_pages(page, 2);
-	gmap_radix_tree_free(&gmap->guest_to_host);
-	gmap_radix_tree_free(&gmap->host_to_guest);
-	down_write(&gmap->mm->mmap_sem);
-	list_del(&gmap->list);
-	up_write(&gmap->mm->mmap_sem);
-	kfree(gmap);
-}
-EXPORT_SYMBOL_GPL(gmap_free);
-
-/**
- * gmap_enable - switch primary space to the guest address space
- * @gmap: pointer to the guest address space structure
- */
-void gmap_enable(struct gmap *gmap)
-{
-	S390_lowcore.gmap = (unsigned long) gmap;
-}
-EXPORT_SYMBOL_GPL(gmap_enable);
-
-/**
- * gmap_disable - switch back to the standard primary address space
- * @gmap: pointer to the guest address space structure
- */
-void gmap_disable(struct gmap *gmap)
-{
-	S390_lowcore.gmap = 0UL;
-}
-EXPORT_SYMBOL_GPL(gmap_disable);
-
-/*
- * gmap_alloc_table is assumed to be called with mmap_sem held
- */
-static int gmap_alloc_table(struct gmap *gmap, unsigned long *table,
-			    unsigned long init, unsigned long gaddr)
-{
-	struct page *page;
-	unsigned long *new;
-
-	/* since we dont free the gmap table until gmap_free we can unlock */
-	page = alloc_pages(GFP_KERNEL, 2);
-	if (!page)
-		return -ENOMEM;
-	new = (unsigned long *) page_to_phys(page);
-	crst_table_init(new, init);
-	spin_lock(&gmap->mm->page_table_lock);
-	if (*table & _REGION_ENTRY_INVALID) {
-		list_add(&page->lru, &gmap->crst_list);
-		*table = (unsigned long) new | _REGION_ENTRY_LENGTH |
-			(*table & _REGION_ENTRY_TYPE_MASK);
-		page->index = gaddr;
-		page = NULL;
-	}
-	spin_unlock(&gmap->mm->page_table_lock);
-	if (page)
-		__free_pages(page, 2);
-	return 0;
-}
-
-/**
- * __gmap_segment_gaddr - find virtual address from segment pointer
- * @entry: pointer to a segment table entry in the guest address space
- *
- * Returns the virtual address in the guest address space for the segment
- */
-static unsigned long __gmap_segment_gaddr(unsigned long *entry)
-{
-	struct page *page;
-	unsigned long offset, mask;
-
-	offset = (unsigned long) entry / sizeof(unsigned long);
-	offset = (offset & (PTRS_PER_PMD - 1)) * PMD_SIZE;
-	mask = ~(PTRS_PER_PMD * sizeof(pmd_t) - 1);
-	page = virt_to_page((void *)((unsigned long) entry & mask));
-	return page->index + offset;
-}
-
-/**
- * __gmap_unlink_by_vmaddr - unlink a single segment via a host address
- * @gmap: pointer to the guest address space structure
- * @vmaddr: address in the host process address space
- *
- * Returns 1 if a TLB flush is required
- */
-static int __gmap_unlink_by_vmaddr(struct gmap *gmap, unsigned long vmaddr)
-{
-	unsigned long *entry;
-	int flush = 0;
-
-	spin_lock(&gmap->guest_table_lock);
-	entry = radix_tree_delete(&gmap->host_to_guest, vmaddr >> PMD_SHIFT);
-	if (entry) {
-		flush = (*entry != _SEGMENT_ENTRY_INVALID);
-		*entry = _SEGMENT_ENTRY_INVALID;
-	}
-	spin_unlock(&gmap->guest_table_lock);
-	return flush;
-}
-
-/**
- * __gmap_unmap_by_gaddr - unmap a single segment via a guest address
- * @gmap: pointer to the guest address space structure
- * @gaddr: address in the guest address space
- *
- * Returns 1 if a TLB flush is required
- */
-static int __gmap_unmap_by_gaddr(struct gmap *gmap, unsigned long gaddr)
-{
-	unsigned long vmaddr;
-
-	vmaddr = (unsigned long) radix_tree_delete(&gmap->guest_to_host,
-						   gaddr >> PMD_SHIFT);
-	return vmaddr ? __gmap_unlink_by_vmaddr(gmap, vmaddr) : 0;
-}
-
-/**
- * gmap_unmap_segment - unmap segment from the guest address space
- * @gmap: pointer to the guest address space structure
- * @to: address in the guest address space
- * @len: length of the memory area to unmap
- *
- * Returns 0 if the unmap succeeded, -EINVAL if not.
- */
-int gmap_unmap_segment(struct gmap *gmap, unsigned long to, unsigned long len)
-{
-	unsigned long off;
-	int flush;
-
-	if ((to | len) & (PMD_SIZE - 1))
-		return -EINVAL;
-	if (len == 0 || to + len < to)
-		return -EINVAL;
-
-	flush = 0;
-	down_write(&gmap->mm->mmap_sem);
-	for (off = 0; off < len; off += PMD_SIZE)
-		flush |= __gmap_unmap_by_gaddr(gmap, to + off);
-	up_write(&gmap->mm->mmap_sem);
-	if (flush)
-		gmap_flush_tlb(gmap);
-	return 0;
-}
-EXPORT_SYMBOL_GPL(gmap_unmap_segment);
-
-/**
- * gmap_mmap_segment - map a segment to the guest address space
- * @gmap: pointer to the guest address space structure
- * @from: source address in the parent address space
- * @to: target address in the guest address space
- * @len: length of the memory area to map
- *
- * Returns 0 if the mmap succeeded, -EINVAL or -ENOMEM if not.
- */
-int gmap_map_segment(struct gmap *gmap, unsigned long from,
-		     unsigned long to, unsigned long len)
-{
-	unsigned long off;
-	int flush;
-
-	if ((from | to | len) & (PMD_SIZE - 1))
-		return -EINVAL;
-	if (len == 0 || from + len < from || to + len < to ||
-	    from + len - 1 > TASK_MAX_SIZE || to + len - 1 > gmap->asce_end)
-		return -EINVAL;
-
-	flush = 0;
-	down_write(&gmap->mm->mmap_sem);
-	for (off = 0; off < len; off += PMD_SIZE) {
-		/* Remove old translation */
-		flush |= __gmap_unmap_by_gaddr(gmap, to + off);
-		/* Store new translation */
-		if (radix_tree_insert(&gmap->guest_to_host,
-				      (to + off) >> PMD_SHIFT,
-				      (void *) from + off))
-			break;
-	}
-	up_write(&gmap->mm->mmap_sem);
-	if (flush)
-		gmap_flush_tlb(gmap);
-	if (off >= len)
-		return 0;
-	gmap_unmap_segment(gmap, to, len);
-	return -ENOMEM;
-}
-EXPORT_SYMBOL_GPL(gmap_map_segment);
-
-/**
- * __gmap_translate - translate a guest address to a user space address
- * @gmap: pointer to guest mapping meta data structure
- * @gaddr: guest address
- *
- * Returns user space address which corresponds to the guest address or
- * -EFAULT if no such mapping exists.
- * This function does not establish potentially missing page table entries.
- * The mmap_sem of the mm that belongs to the address space must be held
- * when this function gets called.
- */
-unsigned long __gmap_translate(struct gmap *gmap, unsigned long gaddr)
-{
-	unsigned long vmaddr;
-
-	vmaddr = (unsigned long)
-		radix_tree_lookup(&gmap->guest_to_host, gaddr >> PMD_SHIFT);
-	return vmaddr ? (vmaddr | (gaddr & ~PMD_MASK)) : -EFAULT;
-}
-EXPORT_SYMBOL_GPL(__gmap_translate);
-
-/**
- * gmap_translate - translate a guest address to a user space address
- * @gmap: pointer to guest mapping meta data structure
- * @gaddr: guest address
- *
- * Returns user space address which corresponds to the guest address or
- * -EFAULT if no such mapping exists.
- * This function does not establish potentially missing page table entries.
- */
-unsigned long gmap_translate(struct gmap *gmap, unsigned long gaddr)
-{
-	unsigned long rc;
-
-	down_read(&gmap->mm->mmap_sem);
-	rc = __gmap_translate(gmap, gaddr);
-	up_read(&gmap->mm->mmap_sem);
-	return rc;
-}
-EXPORT_SYMBOL_GPL(gmap_translate);
-
-/**
- * gmap_unlink - disconnect a page table from the gmap shadow tables
- * @gmap: pointer to guest mapping meta data structure
- * @table: pointer to the host page table
- * @vmaddr: vm address associated with the host page table
- */
-static void gmap_unlink(struct mm_struct *mm, unsigned long *table,
-			unsigned long vmaddr)
-{
-	struct gmap *gmap;
-	int flush;
-
-	list_for_each_entry(gmap, &mm->context.gmap_list, list) {
-		flush = __gmap_unlink_by_vmaddr(gmap, vmaddr);
-		if (flush)
-			gmap_flush_tlb(gmap);
-	}
-}
-
-/**
- * gmap_link - set up shadow page tables to connect a host to a guest address
- * @gmap: pointer to guest mapping meta data structure
- * @gaddr: guest address
- * @vmaddr: vm address
- *
- * Returns 0 on success, -ENOMEM for out of memory conditions, and -EFAULT
- * if the vm address is already mapped to a different guest segment.
- * The mmap_sem of the mm that belongs to the address space must be held
- * when this function gets called.
- */
-int __gmap_link(struct gmap *gmap, unsigned long gaddr, unsigned long vmaddr)
-{
-	struct mm_struct *mm;
-	unsigned long *table;
-	spinlock_t *ptl;
-	pgd_t *pgd;
-	pud_t *pud;
-	pmd_t *pmd;
-	int rc;
-
-	/* Create higher level tables in the gmap page table */
-	table = gmap->table;
-	if ((gmap->asce & _ASCE_TYPE_MASK) >= _ASCE_TYPE_REGION1) {
-		table += (gaddr >> 53) & 0x7ff;
-		if ((*table & _REGION_ENTRY_INVALID) &&
-		    gmap_alloc_table(gmap, table, _REGION2_ENTRY_EMPTY,
-				     gaddr & 0xffe0000000000000UL))
-			return -ENOMEM;
-		table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
-	}
-	if ((gmap->asce & _ASCE_TYPE_MASK) >= _ASCE_TYPE_REGION2) {
-		table += (gaddr >> 42) & 0x7ff;
-		if ((*table & _REGION_ENTRY_INVALID) &&
-		    gmap_alloc_table(gmap, table, _REGION3_ENTRY_EMPTY,
-				     gaddr & 0xfffffc0000000000UL))
-			return -ENOMEM;
-		table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
-	}
-	if ((gmap->asce & _ASCE_TYPE_MASK) >= _ASCE_TYPE_REGION3) {
-		table += (gaddr >> 31) & 0x7ff;
-		if ((*table & _REGION_ENTRY_INVALID) &&
-		    gmap_alloc_table(gmap, table, _SEGMENT_ENTRY_EMPTY,
-				     gaddr & 0xffffffff80000000UL))
-			return -ENOMEM;
-		table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
-	}
-	table += (gaddr >> 20) & 0x7ff;
-	/* Walk the parent mm page table */
-	mm = gmap->mm;
-	pgd = pgd_offset(mm, vmaddr);
-	VM_BUG_ON(pgd_none(*pgd));
-	pud = pud_offset(pgd, vmaddr);
-	VM_BUG_ON(pud_none(*pud));
-	pmd = pmd_offset(pud, vmaddr);
-	VM_BUG_ON(pmd_none(*pmd));
-	/* large pmds cannot yet be handled */
-	if (pmd_large(*pmd))
-		return -EFAULT;
-	/* Link gmap segment table entry location to page table. */
-	rc = radix_tree_preload(GFP_KERNEL);
-	if (rc)
-		return rc;
-	ptl = pmd_lock(mm, pmd);
-	spin_lock(&gmap->guest_table_lock);
-	if (*table == _SEGMENT_ENTRY_INVALID) {
-		rc = radix_tree_insert(&gmap->host_to_guest,
-				       vmaddr >> PMD_SHIFT, table);
-		if (!rc)
-			*table = pmd_val(*pmd);
+	old = *ptep;
+	if (unlikely(pte_val(old) & _PAGE_INVALID))
+		return old;
+	active = (mm == current->active_mm) ? 1 : 0;
+	count = atomic_add_return(0x10000, &mm->context.attach_count);
+	if ((count & 0xffff) <= active) {
+		pte_val(*ptep) |= _PAGE_INVALID;
+		mm->context.flush_mm = 1;
 	} else
-		rc = 0;
-	spin_unlock(&gmap->guest_table_lock);
-	spin_unlock(ptl);
-	radix_tree_preload_end();
-	return rc;
+		__ptep_ipte(addr, ptep);
+	atomic_sub(0x10000, &mm->context.attach_count);
+	return old;
 }
 
-/**
- * gmap_fault - resolve a fault on a guest address
- * @gmap: pointer to guest mapping meta data structure
- * @gaddr: guest address
- * @fault_flags: flags to pass down to handle_mm_fault()
- *
- * Returns 0 on success, -ENOMEM for out of memory conditions, and -EFAULT
- * if the vm address is already mapped to a different guest segment.
- */
-int gmap_fault(struct gmap *gmap, unsigned long gaddr,
-	       unsigned int fault_flags)
+static inline pgste_t pgste_get_lock(pte_t *ptep)
 {
-	unsigned long vmaddr;
-	int rc;
-	bool unlocked;
+	unsigned long new = 0;
+#ifdef CONFIG_PGSTE
+	unsigned long old;
 
-	down_read(&gmap->mm->mmap_sem);
+	preempt_disable();
+	asm(
+		"	lg	%0,%2\n"
+		"0:	lgr	%1,%0\n"
+		"	nihh	%0,0xff7f\n"	/* clear PCL bit in old */
+		"	oihh	%1,0x0080\n"	/* set PCL bit in new */
+		"	csg	%0,%1,%2\n"
+		"	jl	0b\n"
+		: "=&d" (old), "=&d" (new), "=Q" (ptep[PTRS_PER_PTE])
+		: "Q" (ptep[PTRS_PER_PTE]) : "cc", "memory");
+#endif
+	return __pgste(new);
+}
 
-retry:
-	unlocked = false;
-	vmaddr = __gmap_translate(gmap, gaddr);
-	if (IS_ERR_VALUE(vmaddr)) {
-		rc = vmaddr;
-		goto out_up;
-	}
-	if (fixup_user_fault(current, gmap->mm, vmaddr, fault_flags,
-			     &unlocked)) {
-		rc = -EFAULT;
-		goto out_up;
-	}
+static inline void pgste_set_unlock(pte_t *ptep, pgste_t pgste)
+{
+#ifdef CONFIG_PGSTE
+	asm(
+		"	nihh	%1,0xff7f\n"	/* clear PCL bit */
+		"	stg	%1,%0\n"
+		: "=Q" (ptep[PTRS_PER_PTE])
+		: "d" (pgste_val(pgste)), "Q" (ptep[PTRS_PER_PTE])
+		: "cc", "memory");
+	preempt_enable();
+#endif
+}
+
+static inline pgste_t pgste_get(pte_t *ptep)
+{
+	unsigned long pgste = 0;
+#ifdef CONFIG_PGSTE
+	pgste = *(unsigned long *)(ptep + PTRS_PER_PTE);
+#endif
+	return __pgste(pgste);
+}
+
+static inline void pgste_set(pte_t *ptep, pgste_t pgste)
+{
+#ifdef CONFIG_PGSTE
+	*(pgste_t *)(ptep + PTRS_PER_PTE) = pgste;
+#endif
+}
+
+static inline pgste_t pgste_update_all(pte_t pte, pgste_t pgste,
+				       struct mm_struct *mm)
+{
+#ifdef CONFIG_PGSTE
+	unsigned long address, bits, skey;
+
+	if (!mm_use_skey(mm) || pte_val(pte) & _PAGE_INVALID)
+		return pgste;
+	address = pte_val(pte) & PAGE_MASK;
+	skey = (unsigned long) page_get_storage_key(address);
+	bits = skey & (_PAGE_CHANGED | _PAGE_REFERENCED);
+	/* Transfer page changed & referenced bit to guest bits in pgste */
+	pgste_val(pgste) |= bits << 48;		/* GR bit & GC bit */
+	/* Copy page access key and fetch protection bit to pgste */
+	pgste_val(pgste) &= ~(PGSTE_ACC_BITS | PGSTE_FP_BIT);
+	pgste_val(pgste) |= (skey & (_PAGE_ACC_BITS | _PAGE_FP_BIT)) << 56;
+#endif
+	return pgste;
+
+}
+
+static inline void pgste_set_key(pte_t *ptep, pgste_t pgste, pte_t entry,
+				 struct mm_struct *mm)
+{
+#ifdef CONFIG_PGSTE
+	unsigned long address;
+	unsigned long nkey;
+
+	if (!mm_use_skey(mm) || pte_val(entry) & _PAGE_INVALID)
+		return;
+	VM_BUG_ON(!(pte_val(*ptep) & _PAGE_INVALID));
+	address = pte_val(entry) & PAGE_MASK;
 	/*
-	 * In the case that fixup_user_fault unlocked the mmap_sem during
-	 * faultin redo __gmap_translate to not race with a map/unmap_segment.
+	 * Set page access key and fetch protection bit from pgste.
+	 * The guest C/R information is still in the PGSTE, set real
+	 * key C/R to 0.
 	 */
-	if (unlocked)
-		goto retry;
-
-	rc = __gmap_link(gmap, gaddr, vmaddr);
-out_up:
-	up_read(&gmap->mm->mmap_sem);
-	return rc;
-}
-EXPORT_SYMBOL_GPL(gmap_fault);
-
-static void gmap_zap_swap_entry(swp_entry_t entry, struct mm_struct *mm)
-{
-	if (!non_swap_entry(entry))
-		dec_mm_counter(mm, MM_SWAPENTS);
-	else if (is_migration_entry(entry)) {
-		struct page *page = migration_entry_to_page(entry);
-
-		dec_mm_counter(mm, mm_counter(page));
-	}
-	free_swap_and_cache(entry);
+	nkey = (pgste_val(pgste) & (PGSTE_ACC_BITS | PGSTE_FP_BIT)) >> 56;
+	nkey |= (pgste_val(pgste) & (PGSTE_GR_BIT | PGSTE_GC_BIT)) >> 48;
+	page_set_storage_key(address, nkey, 0);
+#endif
 }
 
-/*
- * this function is assumed to be called with mmap_sem held
- */
-void __gmap_zap(struct gmap *gmap, unsigned long gaddr)
+static inline pgste_t pgste_set_pte(pte_t *ptep, pgste_t pgste, pte_t entry)
 {
-	unsigned long vmaddr, ptev, pgstev;
-	pte_t *ptep, pte;
-	spinlock_t *ptl;
-	pgste_t pgste;
-
-	/* Find the vm address for the guest address */
-	vmaddr = (unsigned long) radix_tree_lookup(&gmap->guest_to_host,
-						   gaddr >> PMD_SHIFT);
-	if (!vmaddr)
-		return;
-	vmaddr |= gaddr & ~PMD_MASK;
-	/* Get pointer to the page table entry */
-	ptep = get_locked_pte(gmap->mm, vmaddr, &ptl);
-	if (unlikely(!ptep))
-		return;
-	pte = *ptep;
-	if (!pte_swap(pte))
-		goto out_pte;
-	/* Zap unused and logically-zero pages */
-	pgste = pgste_get_lock(ptep);
-	pgstev = pgste_val(pgste);
-	ptev = pte_val(pte);
-	if (((pgstev & _PGSTE_GPS_USAGE_MASK) == _PGSTE_GPS_USAGE_UNUSED) ||
-	    ((pgstev & _PGSTE_GPS_ZERO) && (ptev & _PAGE_INVALID))) {
-		gmap_zap_swap_entry(pte_to_swp_entry(pte), gmap->mm);
-		pte_clear(gmap->mm, vmaddr, ptep);
-	}
-	pgste_set_unlock(ptep, pgste);
-out_pte:
-	pte_unmap_unlock(ptep, ptl);
-}
-EXPORT_SYMBOL_GPL(__gmap_zap);
-
-void gmap_discard(struct gmap *gmap, unsigned long from, unsigned long to)
-{
-	unsigned long gaddr, vmaddr, size;
-	struct vm_area_struct *vma;
-
-	down_read(&gmap->mm->mmap_sem);
-	for (gaddr = from; gaddr < to;
-	     gaddr = (gaddr + PMD_SIZE) & PMD_MASK) {
-		/* Find the vm address for the guest address */
-		vmaddr = (unsigned long)
-			radix_tree_lookup(&gmap->guest_to_host,
-					  gaddr >> PMD_SHIFT);
-		if (!vmaddr)
-			continue;
-		vmaddr |= gaddr & ~PMD_MASK;
-		/* Find vma in the parent mm */
-		vma = find_vma(gmap->mm, vmaddr);
-		size = min(to - gaddr, PMD_SIZE - (gaddr & ~PMD_MASK));
-		zap_page_range(vma, vmaddr, size, NULL);
-	}
-	up_read(&gmap->mm->mmap_sem);
-}
-EXPORT_SYMBOL_GPL(gmap_discard);
-
-static LIST_HEAD(gmap_notifier_list);
-static DEFINE_SPINLOCK(gmap_notifier_lock);
-
-/**
- * gmap_register_ipte_notifier - register a pte invalidation callback
- * @nb: pointer to the gmap notifier block
- */
-void gmap_register_ipte_notifier(struct gmap_notifier *nb)
-{
-	spin_lock(&gmap_notifier_lock);
-	list_add(&nb->list, &gmap_notifier_list);
-	spin_unlock(&gmap_notifier_lock);
-}
-EXPORT_SYMBOL_GPL(gmap_register_ipte_notifier);
-
-/**
- * gmap_unregister_ipte_notifier - remove a pte invalidation callback
- * @nb: pointer to the gmap notifier block
- */
-void gmap_unregister_ipte_notifier(struct gmap_notifier *nb)
-{
-	spin_lock(&gmap_notifier_lock);
-	list_del_init(&nb->list);
-	spin_unlock(&gmap_notifier_lock);
-}
-EXPORT_SYMBOL_GPL(gmap_unregister_ipte_notifier);
-
-/**
- * gmap_ipte_notify - mark a range of ptes for invalidation notification
- * @gmap: pointer to guest mapping meta data structure
- * @gaddr: virtual address in the guest address space
- * @len: size of area
- *
- * Returns 0 if for each page in the given range a gmap mapping exists and
- * the invalidation notification could be set. If the gmap mapping is missing
- * for one or more pages -EFAULT is returned. If no memory could be allocated
- * -ENOMEM is returned. This function establishes missing page table entries.
- */
-int gmap_ipte_notify(struct gmap *gmap, unsigned long gaddr, unsigned long len)
-{
-	unsigned long addr;
-	spinlock_t *ptl;
-	pte_t *ptep, entry;
-	pgste_t pgste;
-	bool unlocked;
-	int rc = 0;
-
-	if ((gaddr & ~PAGE_MASK) || (len & ~PAGE_MASK))
-		return -EINVAL;
-	down_read(&gmap->mm->mmap_sem);
-	while (len) {
-		unlocked = false;
-		/* Convert gmap address and connect the page tables */
-		addr = __gmap_translate(gmap, gaddr);
-		if (IS_ERR_VALUE(addr)) {
-			rc = addr;
-			break;
+#ifdef CONFIG_PGSTE
+	if ((pte_val(entry) & _PAGE_PRESENT) &&
+	    (pte_val(entry) & _PAGE_WRITE) &&
+	    !(pte_val(entry) & _PAGE_INVALID)) {
+		if (!MACHINE_HAS_ESOP) {
+			/*
+			 * Without enhanced suppression-on-protection force
+			 * the dirty bit on for all writable ptes.
+			 */
+			pte_val(entry) |= _PAGE_DIRTY;
+			pte_val(entry) &= ~_PAGE_PROTECT;
 		}
-		/* Get the page mapped */
-		if (fixup_user_fault(current, gmap->mm, addr, FAULT_FLAG_WRITE,
-				     &unlocked)) {
-			rc = -EFAULT;
-			break;
-		}
-		/* While trying to map mmap_sem got unlocked. Let us retry */
-		if (unlocked)
-			continue;
-		rc = __gmap_link(gmap, gaddr, addr);
-		if (rc)
-			break;
-		/* Walk the process page table, lock and get pte pointer */
-		ptep = get_locked_pte(gmap->mm, addr, &ptl);
-		VM_BUG_ON(!ptep);
-		/* Set notification bit in the pgste of the pte */
-		entry = *ptep;
-		if ((pte_val(entry) & (_PAGE_INVALID | _PAGE_PROTECT)) == 0) {
-			pgste = pgste_get_lock(ptep);
-			pgste_val(pgste) |= PGSTE_IN_BIT;
-			pgste_set_unlock(ptep, pgste);
-			gaddr += PAGE_SIZE;
-			len -= PAGE_SIZE;
-		}
-		pte_unmap_unlock(ptep, ptl);
+		if (!(pte_val(entry) & _PAGE_PROTECT))
+			/* This pte allows write access, set user-dirty */
+			pgste_val(pgste) |= PGSTE_UC_BIT;
 	}
-	up_read(&gmap->mm->mmap_sem);
-	return rc;
+#endif
+	*ptep = entry;
+	return pgste;
 }
-EXPORT_SYMBOL_GPL(gmap_ipte_notify);
 
-/**
- * gmap_do_ipte_notify - call all invalidation callbacks for a specific pte.
- * @mm: pointer to the process mm_struct
- * @addr: virtual address in the process address space
- * @pte: pointer to the page table entry
- *
- * This function is assumed to be called with the page table lock held
- * for the pte to notify.
- */
-void gmap_do_ipte_notify(struct mm_struct *mm, unsigned long vmaddr, pte_t *pte)
+static inline pgste_t pgste_ipte_notify(struct mm_struct *mm,
+					unsigned long addr,
+					pte_t *ptep, pgste_t pgste)
 {
-	unsigned long offset, gaddr;
-	unsigned long *table;
-	struct gmap_notifier *nb;
-	struct gmap *gmap;
-
-	offset = ((unsigned long) pte) & (255 * sizeof(pte_t));
-	offset = offset * (4096 / sizeof(pte_t));
-	spin_lock(&gmap_notifier_lock);
-	list_for_each_entry(gmap, &mm->context.gmap_list, list) {
-		table = radix_tree_lookup(&gmap->host_to_guest,
-					  vmaddr >> PMD_SHIFT);
-		if (!table)
-			continue;
-		gaddr = __gmap_segment_gaddr(table) + offset;
-		list_for_each_entry(nb, &gmap_notifier_list, list)
-			nb->notifier_call(gmap, gaddr);
+#ifdef CONFIG_PGSTE
+	if (pgste_val(pgste) & PGSTE_IN_BIT) {
+		pgste_val(pgste) &= ~PGSTE_IN_BIT;
+		ptep_notify(mm, addr, ptep);
 	}
-	spin_unlock(&gmap_notifier_lock);
+#endif
+	return pgste;
 }
-EXPORT_SYMBOL_GPL(gmap_do_ipte_notify);
 
-int set_guest_storage_key(struct mm_struct *mm, unsigned long addr,
-			  unsigned long key, bool nq)
+static inline pgste_t ptep_xchg_start(struct mm_struct *mm,
+				      unsigned long addr, pte_t *ptep)
 {
-	spinlock_t *ptl;
-	pgste_t old, new;
-	pte_t *ptep;
-	bool unlocked;
+	pgste_t pgste = __pgste(0);
 
-	down_read(&mm->mmap_sem);
-retry:
-	unlocked = false;
-	ptep = get_locked_pte(mm, addr, &ptl);
-	if (unlikely(!ptep)) {
-		up_read(&mm->mmap_sem);
-		return -EFAULT;
+	if (mm_has_pgste(mm)) {
+		pgste = pgste_get_lock(ptep);
+		pgste = pgste_ipte_notify(mm, addr, ptep, pgste);
 	}
-	if (!(pte_val(*ptep) & _PAGE_INVALID) &&
-	     (pte_val(*ptep) & _PAGE_PROTECT)) {
-		pte_unmap_unlock(ptep, ptl);
-		/*
-		 * We do not really care about unlocked. We will retry either
-		 * way. But this allows fixup_user_fault to enable userfaultfd.
-		 */
-		if (fixup_user_fault(current, mm, addr, FAULT_FLAG_WRITE,
-				     &unlocked)) {
-			up_read(&mm->mmap_sem);
-			return -EFAULT;
+	return pgste;
+}
+
+static inline void ptep_xchg_commit(struct mm_struct *mm,
+				    unsigned long addr, pte_t *ptep,
+				    pgste_t pgste, pte_t old, pte_t new)
+{
+	if (mm_has_pgste(mm)) {
+		if (pte_val(old) & _PAGE_INVALID)
+			pgste_set_key(ptep, pgste, new, mm);
+		if (pte_val(new) & _PAGE_INVALID) {
+			pgste = pgste_update_all(old, pgste, mm);
+			if ((pgste_val(pgste) & _PGSTE_GPS_USAGE_MASK) ==
+			    _PGSTE_GPS_USAGE_UNUSED)
+				pte_val(old) |= _PAGE_UNUSED;
 		}
-		goto retry;
-	}
-
-	new = old = pgste_get_lock(ptep);
-	pgste_val(new) &= ~(PGSTE_GR_BIT | PGSTE_GC_BIT |
-			    PGSTE_ACC_BITS | PGSTE_FP_BIT);
-	pgste_val(new) |= (key & (_PAGE_CHANGED | _PAGE_REFERENCED)) << 48;
-	pgste_val(new) |= (key & (_PAGE_ACC_BITS | _PAGE_FP_BIT)) << 56;
-	if (!(pte_val(*ptep) & _PAGE_INVALID)) {
-		unsigned long address, bits, skey;
-
-		address = pte_val(*ptep) & PAGE_MASK;
-		skey = (unsigned long) page_get_storage_key(address);
-		bits = skey & (_PAGE_CHANGED | _PAGE_REFERENCED);
-		skey = key & (_PAGE_ACC_BITS | _PAGE_FP_BIT);
-		/* Set storage key ACC and FP */
-		page_set_storage_key(address, skey, !nq);
-		/* Merge host changed & referenced into pgste  */
-		pgste_val(new) |= bits << 52;
-	}
-	/* changing the guest storage key is considered a change of the page */
-	if ((pgste_val(new) ^ pgste_val(old)) &
-	    (PGSTE_ACC_BITS | PGSTE_FP_BIT | PGSTE_GR_BIT | PGSTE_GC_BIT))
-		pgste_val(new) |= PGSTE_UC_BIT;
-
-	pgste_set_unlock(ptep, new);
-	pte_unmap_unlock(ptep, ptl);
-	up_read(&mm->mmap_sem);
-	return 0;
-}
-EXPORT_SYMBOL(set_guest_storage_key);
-
-unsigned long get_guest_storage_key(struct mm_struct *mm, unsigned long addr)
-{
-	spinlock_t *ptl;
-	pgste_t pgste;
-	pte_t *ptep;
-	uint64_t physaddr;
-	unsigned long key = 0;
-
-	down_read(&mm->mmap_sem);
-	ptep = get_locked_pte(mm, addr, &ptl);
-	if (unlikely(!ptep)) {
-		up_read(&mm->mmap_sem);
-		return -EFAULT;
-	}
-	pgste = pgste_get_lock(ptep);
-
-	if (pte_val(*ptep) & _PAGE_INVALID) {
-		key |= (pgste_val(pgste) & PGSTE_ACC_BITS) >> 56;
-		key |= (pgste_val(pgste) & PGSTE_FP_BIT) >> 56;
-		key |= (pgste_val(pgste) & PGSTE_GR_BIT) >> 48;
-		key |= (pgste_val(pgste) & PGSTE_GC_BIT) >> 48;
+		pgste = pgste_set_pte(ptep, pgste, new);
+		pgste_set_unlock(ptep, pgste);
 	} else {
-		physaddr = pte_val(*ptep) & PAGE_MASK;
-		key = page_get_storage_key(physaddr);
-
-		/* Reflect guest's logical view, not physical */
-		if (pgste_val(pgste) & PGSTE_GR_BIT)
-			key |= _PAGE_REFERENCED;
-		if (pgste_val(pgste) & PGSTE_GC_BIT)
-			key |= _PAGE_CHANGED;
+		*ptep = new;
 	}
-
-	pgste_set_unlock(ptep, pgste);
-	pte_unmap_unlock(ptep, ptl);
-	up_read(&mm->mmap_sem);
-	return key;
-}
-EXPORT_SYMBOL(get_guest_storage_key);
-
-static int page_table_allocate_pgste_min = 0;
-static int page_table_allocate_pgste_max = 1;
-int page_table_allocate_pgste = 0;
-EXPORT_SYMBOL(page_table_allocate_pgste);
-
-static struct ctl_table page_table_sysctl[] = {
-	{
-		.procname	= "allocate_pgste",
-		.data		= &page_table_allocate_pgste,
-		.maxlen		= sizeof(int),
-		.mode		= S_IRUGO | S_IWUSR,
-		.proc_handler	= proc_dointvec,
-		.extra1		= &page_table_allocate_pgste_min,
-		.extra2		= &page_table_allocate_pgste_max,
-	},
-	{ }
-};
-
-static struct ctl_table page_table_sysctl_dir[] = {
-	{
-		.procname	= "vm",
-		.maxlen		= 0,
-		.mode		= 0555,
-		.child		= page_table_sysctl,
-	},
-	{ }
-};
-
-static int __init page_table_register_sysctl(void)
-{
-	return register_sysctl_table(page_table_sysctl_dir) ? 0 : -ENOMEM;
-}
-__initcall(page_table_register_sysctl);
-
-#else /* CONFIG_PGSTE */
-
-static inline void gmap_unlink(struct mm_struct *mm, unsigned long *table,
-			unsigned long vmaddr)
-{
 }
 
-#endif /* CONFIG_PGSTE */
-
-static inline unsigned int atomic_xor_bits(atomic_t *v, unsigned int bits)
+pte_t ptep_xchg_direct(struct mm_struct *mm, unsigned long addr,
+		       pte_t *ptep, pte_t new)
 {
-	unsigned int old, new;
+	pgste_t pgste;
+	pte_t old;
 
-	do {
-		old = atomic_read(v);
-		new = old ^ bits;
-	} while (atomic_cmpxchg(v, old, new) != old);
-	return new;
+	pgste = ptep_xchg_start(mm, addr, ptep);
+	old = ptep_flush_direct(mm, addr, ptep);
+	ptep_xchg_commit(mm, addr, ptep, pgste, old, new);
+	return old;
 }
+EXPORT_SYMBOL(ptep_xchg_direct);
 
-/*
- * page table entry allocation/free routines.
- */
-unsigned long *page_table_alloc(struct mm_struct *mm)
+pte_t ptep_xchg_lazy(struct mm_struct *mm, unsigned long addr,
+		     pte_t *ptep, pte_t new)
 {
-	unsigned long *table;
-	struct page *page;
-	unsigned int mask, bit;
+	pgste_t pgste;
+	pte_t old;
 
-	/* Try to get a fragment of a 4K page as a 2K page table */
-	if (!mm_alloc_pgste(mm)) {
-		table = NULL;
-		spin_lock_bh(&mm->context.list_lock);
-		if (!list_empty(&mm->context.pgtable_list)) {
-			page = list_first_entry(&mm->context.pgtable_list,
-						struct page, lru);
-			mask = atomic_read(&page->_mapcount);
-			mask = (mask | (mask >> 4)) & 3;
-			if (mask != 3) {
-				table = (unsigned long *) page_to_phys(page);
-				bit = mask & 1;		/* =1 -> second 2K */
-				if (bit)
-					table += PTRS_PER_PTE;
-				atomic_xor_bits(&page->_mapcount, 1U << bit);
-				list_del(&page->lru);
-			}
-		}
-		spin_unlock_bh(&mm->context.list_lock);
-		if (table)
-			return table;
+	pgste = ptep_xchg_start(mm, addr, ptep);
+	old = ptep_flush_lazy(mm, addr, ptep);
+	ptep_xchg_commit(mm, addr, ptep, pgste, old, new);
+	return old;
+}
+EXPORT_SYMBOL(ptep_xchg_lazy);
+
+pte_t ptep_modify_prot_start(struct mm_struct *mm, unsigned long addr,
+			     pte_t *ptep)
+{
+	pgste_t pgste;
+	pte_t old;
+
+	pgste = ptep_xchg_start(mm, addr, ptep);
+	old = ptep_flush_lazy(mm, addr, ptep);
+	if (mm_has_pgste(mm)) {
+		pgste = pgste_update_all(old, pgste, mm);
+		pgste_set(ptep, pgste);
 	}
-	/* Allocate a fresh page */
-	page = alloc_page(GFP_KERNEL|__GFP_REPEAT);
-	if (!page)
-		return NULL;
-	if (!pgtable_page_ctor(page)) {
-		__free_page(page);
-		return NULL;
-	}
-	/* Initialize page table */
-	table = (unsigned long *) page_to_phys(page);
-	if (mm_alloc_pgste(mm)) {
-		/* Return 4K page table with PGSTEs */
-		atomic_set(&page->_mapcount, 3);
-		clear_table(table, _PAGE_INVALID, PAGE_SIZE/2);
-		clear_table(table + PTRS_PER_PTE, 0, PAGE_SIZE/2);
+	return old;
+}
+EXPORT_SYMBOL(ptep_modify_prot_start);
+
+void ptep_modify_prot_commit(struct mm_struct *mm, unsigned long addr,
+			     pte_t *ptep, pte_t pte)
+{
+	pgste_t pgste;
+
+	if (mm_has_pgste(mm)) {
+		pgste = pgste_get(ptep);
+		pgste_set_key(ptep, pgste, pte, mm);
+		pgste = pgste_set_pte(ptep, pgste, pte);
+		pgste_set_unlock(ptep, pgste);
 	} else {
-		/* Return the first 2K fragment of the page */
-		atomic_set(&page->_mapcount, 1);
-		clear_table(table, _PAGE_INVALID, PAGE_SIZE);
-		spin_lock_bh(&mm->context.list_lock);
-		list_add(&page->lru, &mm->context.pgtable_list);
-		spin_unlock_bh(&mm->context.list_lock);
+		*ptep = pte;
 	}
-	return table;
 }
+EXPORT_SYMBOL(ptep_modify_prot_commit);
 
-void page_table_free(struct mm_struct *mm, unsigned long *table)
+static inline pmd_t pmdp_flush_direct(struct mm_struct *mm,
+				      unsigned long addr, pmd_t *pmdp)
 {
-	struct page *page;
-	unsigned int bit, mask;
+	int active, count;
+	pmd_t old;
 
-	page = pfn_to_page(__pa(table) >> PAGE_SHIFT);
-	if (!mm_alloc_pgste(mm)) {
-		/* Free 2K page table fragment of a 4K page */
-		bit = (__pa(table) & ~PAGE_MASK)/(PTRS_PER_PTE*sizeof(pte_t));
-		spin_lock_bh(&mm->context.list_lock);
-		mask = atomic_xor_bits(&page->_mapcount, 1U << bit);
-		if (mask & 3)
-			list_add(&page->lru, &mm->context.pgtable_list);
-		else
-			list_del(&page->lru);
-		spin_unlock_bh(&mm->context.list_lock);
-		if (mask != 0)
-			return;
+	old = *pmdp;
+	if (pmd_val(old) & _SEGMENT_ENTRY_INVALID)
+		return old;
+	if (!MACHINE_HAS_IDTE) {
+		__pmdp_csp(pmdp);
+		return old;
 	}
-
-	pgtable_page_dtor(page);
-	atomic_set(&page->_mapcount, -1);
-	__free_page(page);
-}
-
-void page_table_free_rcu(struct mmu_gather *tlb, unsigned long *table,
-			 unsigned long vmaddr)
-{
-	struct mm_struct *mm;
-	struct page *page;
-	unsigned int bit, mask;
-
-	mm = tlb->mm;
-	page = pfn_to_page(__pa(table) >> PAGE_SHIFT);
-	if (mm_alloc_pgste(mm)) {
-		gmap_unlink(mm, table, vmaddr);
-		table = (unsigned long *) (__pa(table) | 3);
-		tlb_remove_table(tlb, table);
-		return;
-	}
-	bit = (__pa(table) & ~PAGE_MASK) / (PTRS_PER_PTE*sizeof(pte_t));
-	spin_lock_bh(&mm->context.list_lock);
-	mask = atomic_xor_bits(&page->_mapcount, 0x11U << bit);
-	if (mask & 3)
-		list_add_tail(&page->lru, &mm->context.pgtable_list);
+	active = (mm == current->active_mm) ? 1 : 0;
+	count = atomic_add_return(0x10000, &mm->context.attach_count);
+	if (MACHINE_HAS_TLB_LC && (count & 0xffff) <= active &&
+	    cpumask_equal(mm_cpumask(mm), cpumask_of(smp_processor_id())))
+		__pmdp_idte_local(addr, pmdp);
 	else
-		list_del(&page->lru);
-	spin_unlock_bh(&mm->context.list_lock);
-	table = (unsigned long *) (__pa(table) | (1U << bit));
-	tlb_remove_table(tlb, table);
+		__pmdp_idte(addr, pmdp);
+	atomic_sub(0x10000, &mm->context.attach_count);
+	return old;
 }
 
-static void __tlb_remove_table(void *_table)
+static inline pmd_t pmdp_flush_lazy(struct mm_struct *mm,
+				    unsigned long addr, pmd_t *pmdp)
 {
-	unsigned int mask = (unsigned long) _table & 3;
-	void *table = (void *)((unsigned long) _table ^ mask);
-	struct page *page = pfn_to_page(__pa(table) >> PAGE_SHIFT);
+	int active, count;
+	pmd_t old;
 
-	switch (mask) {
-	case 0:		/* pmd or pud */
-		free_pages((unsigned long) table, 2);
-		break;
-	case 1:		/* lower 2K of a 4K page table */
-	case 2:		/* higher 2K of a 4K page table */
-		if (atomic_xor_bits(&page->_mapcount, mask << 4) != 0)
-			break;
-		/* fallthrough */
-	case 3:		/* 4K page table with pgstes */
-		pgtable_page_dtor(page);
-		atomic_set(&page->_mapcount, -1);
-		__free_page(page);
-		break;
-	}
+	old = *pmdp;
+	if (pmd_val(old) & _SEGMENT_ENTRY_INVALID)
+		return old;
+	active = (mm == current->active_mm) ? 1 : 0;
+	count = atomic_add_return(0x10000, &mm->context.attach_count);
+	if ((count & 0xffff) <= active) {
+		pmd_val(*pmdp) |= _SEGMENT_ENTRY_INVALID;
+		mm->context.flush_mm = 1;
+	} else if (MACHINE_HAS_IDTE)
+		__pmdp_idte(addr, pmdp);
+	else
+		__pmdp_csp(pmdp);
+	atomic_sub(0x10000, &mm->context.attach_count);
+	return old;
 }
 
-static void tlb_remove_table_smp_sync(void *arg)
+pmd_t pmdp_xchg_direct(struct mm_struct *mm, unsigned long addr,
+		       pmd_t *pmdp, pmd_t new)
 {
-	/* Simply deliver the interrupt */
-}
+	pmd_t old;
 
-static void tlb_remove_table_one(void *table)
+	old = pmdp_flush_direct(mm, addr, pmdp);
+	*pmdp = new;
+	return old;
+}
+EXPORT_SYMBOL(pmdp_xchg_direct);
+
+pmd_t pmdp_xchg_lazy(struct mm_struct *mm, unsigned long addr,
+		     pmd_t *pmdp, pmd_t new)
 {
-	/*
-	 * This isn't an RCU grace period and hence the page-tables cannot be
-	 * assumed to be actually RCU-freed.
-	 *
-	 * It is however sufficient for software page-table walkers that rely
-	 * on IRQ disabling. See the comment near struct mmu_table_batch.
-	 */
-	smp_call_function(tlb_remove_table_smp_sync, NULL, 1);
-	__tlb_remove_table(table);
+	pmd_t old;
+
+	old = pmdp_flush_lazy(mm, addr, pmdp);
+	*pmdp = new;
+	return old;
 }
-
-static void tlb_remove_table_rcu(struct rcu_head *head)
-{
-	struct mmu_table_batch *batch;
-	int i;
-
-	batch = container_of(head, struct mmu_table_batch, rcu);
-
-	for (i = 0; i < batch->nr; i++)
-		__tlb_remove_table(batch->tables[i]);
-
-	free_page((unsigned long)batch);
-}
-
-void tlb_table_flush(struct mmu_gather *tlb)
-{
-	struct mmu_table_batch **batch = &tlb->batch;
-
-	if (*batch) {
-		call_rcu_sched(&(*batch)->rcu, tlb_remove_table_rcu);
-		*batch = NULL;
-	}
-}
-
-void tlb_remove_table(struct mmu_gather *tlb, void *table)
-{
-	struct mmu_table_batch **batch = &tlb->batch;
-
-	tlb->mm->context.flush_mm = 1;
-	if (*batch == NULL) {
-		*batch = (struct mmu_table_batch *)
-			__get_free_page(GFP_NOWAIT | __GFP_NOWARN);
-		if (*batch == NULL) {
-			__tlb_flush_mm_lazy(tlb->mm);
-			tlb_remove_table_one(table);
-			return;
-		}
-		(*batch)->nr = 0;
-	}
-	(*batch)->tables[(*batch)->nr++] = table;
-	if ((*batch)->nr == MAX_TABLE_BATCH)
-		tlb_flush_mmu(tlb);
-}
+EXPORT_SYMBOL(pmdp_xchg_lazy);
 
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
-static inline void thp_split_vma(struct vm_area_struct *vma)
-{
-	unsigned long addr;
-
-	for (addr = vma->vm_start; addr < vma->vm_end; addr += PAGE_SIZE)
-		follow_page(vma, addr, FOLL_SPLIT);
-}
-
-static inline void thp_split_mm(struct mm_struct *mm)
-{
-	struct vm_area_struct *vma;
-
-	for (vma = mm->mmap; vma != NULL; vma = vma->vm_next) {
-		thp_split_vma(vma);
-		vma->vm_flags &= ~VM_HUGEPAGE;
-		vma->vm_flags |= VM_NOHUGEPAGE;
-	}
-	mm->def_flags |= VM_NOHUGEPAGE;
-}
-#else
-static inline void thp_split_mm(struct mm_struct *mm)
-{
-}
-#endif /* CONFIG_TRANSPARENT_HUGEPAGE */
-
-/*
- * switch on pgstes for its userspace process (for kvm)
- */
-int s390_enable_sie(void)
-{
-	struct mm_struct *mm = current->mm;
-
-	/* Do we have pgstes? if yes, we are done */
-	if (mm_has_pgste(mm))
-		return 0;
-	/* Fail if the page tables are 2K */
-	if (!mm_alloc_pgste(mm))
-		return -EINVAL;
-	down_write(&mm->mmap_sem);
-	mm->context.has_pgste = 1;
-	/* split thp mappings and disable thp for future mappings */
-	thp_split_mm(mm);
-	up_write(&mm->mmap_sem);
-	return 0;
-}
-EXPORT_SYMBOL_GPL(s390_enable_sie);
-
-/*
- * Enable storage key handling from now on and initialize the storage
- * keys with the default key.
- */
-static int __s390_enable_skey(pte_t *pte, unsigned long addr,
-			      unsigned long next, struct mm_walk *walk)
-{
-	unsigned long ptev;
-	pgste_t pgste;
-
-	pgste = pgste_get_lock(pte);
-	/*
-	 * Remove all zero page mappings,
-	 * after establishing a policy to forbid zero page mappings
-	 * following faults for that page will get fresh anonymous pages
-	 */
-	if (is_zero_pfn(pte_pfn(*pte))) {
-		ptep_flush_direct(walk->mm, addr, pte);
-		pte_val(*pte) = _PAGE_INVALID;
-	}
-	/* Clear storage key */
-	pgste_val(pgste) &= ~(PGSTE_ACC_BITS | PGSTE_FP_BIT |
-			      PGSTE_GR_BIT | PGSTE_GC_BIT);
-	ptev = pte_val(*pte);
-	if (!(ptev & _PAGE_INVALID) && (ptev & _PAGE_WRITE))
-		page_set_storage_key(ptev & PAGE_MASK, PAGE_DEFAULT_KEY, 1);
-	pgste_set_unlock(pte, pgste);
-	return 0;
-}
-
-int s390_enable_skey(void)
-{
-	struct mm_walk walk = { .pte_entry = __s390_enable_skey };
-	struct mm_struct *mm = current->mm;
-	struct vm_area_struct *vma;
-	int rc = 0;
-
-	down_write(&mm->mmap_sem);
-	if (mm_use_skey(mm))
-		goto out_up;
-
-	mm->context.use_skey = 1;
-	for (vma = mm->mmap; vma; vma = vma->vm_next) {
-		if (ksm_madvise(vma, vma->vm_start, vma->vm_end,
-				MADV_UNMERGEABLE, &vma->vm_flags)) {
-			mm->context.use_skey = 0;
-			rc = -ENOMEM;
-			goto out_up;
-		}
-	}
-	mm->def_flags &= ~VM_MERGEABLE;
-
-	walk.mm = mm;
-	walk_page_range(0, TASK_SIZE, &walk);
-
-out_up:
-	up_write(&mm->mmap_sem);
-	return rc;
-}
-EXPORT_SYMBOL_GPL(s390_enable_skey);
-
-/*
- * Reset CMMA state, make all pages stable again.
- */
-static int __s390_reset_cmma(pte_t *pte, unsigned long addr,
-			     unsigned long next, struct mm_walk *walk)
-{
-	pgste_t pgste;
-
-	pgste = pgste_get_lock(pte);
-	pgste_val(pgste) &= ~_PGSTE_GPS_USAGE_MASK;
-	pgste_set_unlock(pte, pgste);
-	return 0;
-}
-
-void s390_reset_cmma(struct mm_struct *mm)
-{
-	struct mm_walk walk = { .pte_entry = __s390_reset_cmma };
-
-	down_write(&mm->mmap_sem);
-	walk.mm = mm;
-	walk_page_range(0, TASK_SIZE, &walk);
-	up_write(&mm->mmap_sem);
-}
-EXPORT_SYMBOL_GPL(s390_reset_cmma);
-
-/*
- * Test and reset if a guest page is dirty
- */
-bool gmap_test_and_clear_dirty(unsigned long address, struct gmap *gmap)
-{
-	pte_t *pte;
-	spinlock_t *ptl;
-	bool dirty = false;
-
-	pte = get_locked_pte(gmap->mm, address, &ptl);
-	if (unlikely(!pte))
-		return false;
-
-	if (ptep_test_and_clear_user_dirty(gmap->mm, address, pte))
-		dirty = true;
-
-	spin_unlock(ptl);
-	return dirty;
-}
-EXPORT_SYMBOL_GPL(gmap_test_and_clear_dirty);
-
-#ifdef CONFIG_TRANSPARENT_HUGEPAGE
-int pmdp_clear_flush_young(struct vm_area_struct *vma, unsigned long address,
-			   pmd_t *pmdp)
-{
-	VM_BUG_ON(address & ~HPAGE_PMD_MASK);
-	/* No need to flush TLB
-	 * On s390 reference bits are in storage key and never in TLB */
-	return pmdp_test_and_clear_young(vma, address, pmdp);
-}
-
-int pmdp_set_access_flags(struct vm_area_struct *vma,
-			  unsigned long address, pmd_t *pmdp,
-			  pmd_t entry, int dirty)
-{
-	VM_BUG_ON(address & ~HPAGE_PMD_MASK);
-
-	entry = pmd_mkyoung(entry);
-	if (dirty)
-		entry = pmd_mkdirty(entry);
-	if (pmd_same(*pmdp, entry))
-		return 0;
-	pmdp_invalidate(vma, address, pmdp);
-	set_pmd_at(vma->vm_mm, address, pmdp, entry);
-	return 1;
-}
-
 void pgtable_trans_huge_deposit(struct mm_struct *mm, pmd_t *pmdp,
 				pgtable_t pgtable)
 {
···
 	return pgtable;
 }
 #endif /* CONFIG_TRANSPARENT_HUGEPAGE */
+
+#ifdef CONFIG_PGSTE
+void ptep_set_pte_at(struct mm_struct *mm, unsigned long addr,
+		     pte_t *ptep, pte_t entry)
+{
+	pgste_t pgste;
+
+	/* the mm_has_pgste() check is done in set_pte_at() */
+	pgste = pgste_get_lock(ptep);
+	pgste_val(pgste) &= ~_PGSTE_GPS_ZERO;
+	pgste_set_key(ptep, pgste, entry, mm);
+	pgste = pgste_set_pte(ptep, pgste, entry);
+	pgste_set_unlock(ptep, pgste);
+}
+
+void ptep_set_notify(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
+{
+	pgste_t pgste;
+
+	pgste = pgste_get_lock(ptep);
+	pgste_val(pgste) |= PGSTE_IN_BIT;
+	pgste_set_unlock(ptep, pgste);
+}
+
+static void ptep_zap_swap_entry(struct mm_struct *mm, swp_entry_t entry)
+{
+	if (!non_swap_entry(entry))
+		dec_mm_counter(mm, MM_SWAPENTS);
+	else if (is_migration_entry(entry)) {
+		struct page *page = migration_entry_to_page(entry);
+
+		dec_mm_counter(mm, mm_counter(page));
+	}
+	free_swap_and_cache(entry);
+}
+
+void ptep_zap_unused(struct mm_struct *mm, unsigned long addr,
+		     pte_t *ptep, int reset)
+{
+	unsigned long pgstev;
+	pgste_t pgste;
+	pte_t pte;
+
+	/* Zap unused and logically-zero pages */
+	pgste = pgste_get_lock(ptep);
+	pgstev = pgste_val(pgste);
+	pte = *ptep;
+	if (pte_swap(pte) &&
+	    ((pgstev & _PGSTE_GPS_USAGE_MASK) == _PGSTE_GPS_USAGE_UNUSED ||
+	     (pgstev & _PGSTE_GPS_ZERO))) {
+		ptep_zap_swap_entry(mm, pte_to_swp_entry(pte));
+		pte_clear(mm, addr, ptep);
+	}
+	if (reset)
+		pgste_val(pgste) &= ~_PGSTE_GPS_USAGE_MASK;
+	pgste_set_unlock(ptep, pgste);
+}
+
+void ptep_zap_key(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
+{
+	unsigned long ptev;
+	pgste_t pgste;
+
+	/* Clear storage key */
+	pgste = pgste_get_lock(ptep);
+	pgste_val(pgste) &= ~(PGSTE_ACC_BITS | PGSTE_FP_BIT |
+			      PGSTE_GR_BIT | PGSTE_GC_BIT);
+	ptev = pte_val(*ptep);
+	if (!(ptev & _PAGE_INVALID) && (ptev & _PAGE_WRITE))
+		page_set_storage_key(ptev & PAGE_MASK, PAGE_DEFAULT_KEY, 1);
+	pgste_set_unlock(ptep, pgste);
+}
+
+/*
+ * Test and reset if a guest page is dirty
+ */
+bool test_and_clear_guest_dirty(struct mm_struct *mm, unsigned long addr)
+{
+	spinlock_t *ptl;
+	pgste_t pgste;
+	pte_t *ptep;
+	pte_t pte;
+	bool dirty;
+
+	ptep = get_locked_pte(mm, addr, &ptl);
+	if (unlikely(!ptep))
+		return false;
+
+	pgste = pgste_get_lock(ptep);
+	dirty = !!(pgste_val(pgste) & PGSTE_UC_BIT);
+	pgste_val(pgste) &= ~PGSTE_UC_BIT;
+	pte = *ptep;
+	if (dirty && (pte_val(pte) & _PAGE_PRESENT)) {
+		pgste = pgste_ipte_notify(mm, addr, ptep, pgste);
+		__ptep_ipte(addr, ptep);
+		if (MACHINE_HAS_ESOP || !(pte_val(pte) & _PAGE_WRITE))
+			pte_val(pte) |= _PAGE_PROTECT;
+		else
+			pte_val(pte) |= _PAGE_INVALID;
+		*ptep = pte;
+	}
+	pgste_set_unlock(ptep, pgste);
+
+	spin_unlock(ptl);
+	return dirty;
+}
+EXPORT_SYMBOL_GPL(test_and_clear_guest_dirty);
+
+int set_guest_storage_key(struct mm_struct *mm, unsigned long addr,
+			  unsigned char key, bool nq)
+{
+	unsigned long keyul;
+	spinlock_t *ptl;
+	pgste_t old, new;
+	pte_t *ptep;
+
+	down_read(&mm->mmap_sem);
+	ptep = get_locked_pte(mm, addr, &ptl);
+	if (unlikely(!ptep)) {
+		up_read(&mm->mmap_sem);
+		return -EFAULT;
+	}
+
+	new = old = pgste_get_lock(ptep);
+	pgste_val(new) &= ~(PGSTE_GR_BIT | PGSTE_GC_BIT |
+			    PGSTE_ACC_BITS | PGSTE_FP_BIT);
+	keyul = (unsigned long) key;
+	pgste_val(new) |= (keyul & (_PAGE_CHANGED | _PAGE_REFERENCED)) << 48;
+	pgste_val(new) |= (keyul & (_PAGE_ACC_BITS | _PAGE_FP_BIT)) << 56;
+	if (!(pte_val(*ptep) & _PAGE_INVALID)) {
+		unsigned long address, bits, skey;
+
+		address = pte_val(*ptep) & PAGE_MASK;
+		skey = (unsigned long) page_get_storage_key(address);
+		bits = skey & (_PAGE_CHANGED | _PAGE_REFERENCED);
+		skey = key & (_PAGE_ACC_BITS | _PAGE_FP_BIT);
+		/* Set storage key ACC and FP */
+		page_set_storage_key(address, skey, !nq);
+		/* Merge host changed & referenced into pgste  */
+		pgste_val(new) |= bits << 52;
+	}
+	/* changing the guest storage key is considered a change of the page */
+	if ((pgste_val(new) ^ pgste_val(old)) &
+	    (PGSTE_ACC_BITS | PGSTE_FP_BIT | PGSTE_GR_BIT | PGSTE_GC_BIT))
+		pgste_val(new) |= PGSTE_UC_BIT;
+
+	pgste_set_unlock(ptep, new);
+	pte_unmap_unlock(ptep, ptl);
+	up_read(&mm->mmap_sem);
+	return 0;
+}
+EXPORT_SYMBOL(set_guest_storage_key);
+
+unsigned char get_guest_storage_key(struct mm_struct *mm, unsigned long addr)
+{
+	unsigned char key;
+	spinlock_t *ptl;
+	pgste_t pgste;
+	pte_t *ptep;
+
+	down_read(&mm->mmap_sem);
+	ptep = get_locked_pte(mm, addr, &ptl);
+	if (unlikely(!ptep)) {
+		up_read(&mm->mmap_sem);
+		return -EFAULT;
+	}
+	pgste = pgste_get_lock(ptep);
+
+	if (pte_val(*ptep) & _PAGE_INVALID) {
+		key  = (pgste_val(pgste) & PGSTE_ACC_BITS) >> 56;
+		key |= (pgste_val(pgste) & PGSTE_FP_BIT) >> 56;
+		key |= (pgste_val(pgste) & PGSTE_GR_BIT) >> 48;
+		key |= (pgste_val(pgste) & PGSTE_GC_BIT) >> 48;
+	} else {
+		key = page_get_storage_key(pte_val(*ptep) & PAGE_MASK);
+
+		/* Reflect guest's logical view, not physical */
+		if (pgste_val(pgste) & PGSTE_GR_BIT)
+			key |= _PAGE_REFERENCED;
+		if (pgste_val(pgste) & PGSTE_GC_BIT)
+			key |= _PAGE_CHANGED;
+	}
+
+	pgste_set_unlock(ptep, pgste);
+	pte_unmap_unlock(ptep, ptl);
+	up_read(&mm->mmap_sem);
+	return key;
+}
+EXPORT_SYMBOL(get_guest_storage_key);
+#endif

arch/s390/oprofile/Makefile

···
 		oprofilefs.o oprofile_stats.o  \
 		timer_int.o )
 
-oprofile-y :=	$(DRIVER_OBJS) init.o backtrace.o
+oprofile-y :=	$(DRIVER_OBJS) init.o
 oprofile-y +=	hwsampler.o

arch/s390/oprofile/backtrace.c

···
-/*
- * S390 Version
- *   Copyright IBM Corp. 2005
- *   Author(s): Andreas Krebbel <[e-mail]>
- */
-
-#include <linux/oprofile.h>
-
-#include <asm/processor.h> /* for struct stack_frame */
-
-static unsigned long
-__show_trace(unsigned int *depth, unsigned long sp,
-	     unsigned long low, unsigned long high)
-{
-	struct stack_frame *sf;
-	struct pt_regs *regs;
-
-	while (*depth) {
-		if (sp < low || sp > high - sizeof(*sf))
-			return sp;
-		sf = (struct stack_frame *) sp;
-		(*depth)--;
-		oprofile_add_trace(sf->gprs[8]);
-
-		/* Follow the backchain.  */
-		while (*depth) {
-			low = sp;
-			sp = sf->back_chain;
-			if (!sp)
-				break;
-			if (sp <= low || sp > high - sizeof(*sf))
-				return sp;
-			sf = (struct stack_frame *) sp;
-			(*depth)--;
-			oprofile_add_trace(sf->gprs[8]);
-
-		}
-
-		if (*depth == 0)
-			break;
-
-		/* Zero backchain detected, check for interrupt frame.  */
-		sp = (unsigned long) (sf + 1);
-		if (sp <= low || sp > high - sizeof(*regs))
-			return sp;
-		regs = (struct pt_regs *) sp;
-		(*depth)--;
-		oprofile_add_trace(sf->gprs[8]);
-		low = sp;
-		sp = regs->gprs[15];
-	}
-	return sp;
-}
-
-void s390_backtrace(struct pt_regs * const regs, unsigned int depth)
-{
-	unsigned long head, frame_size;
-	struct stack_frame* head_sf;
-
-	if (user_mode(regs))
-		return;
-
-	frame_size = STACK_FRAME_OVERHEAD + sizeof(struct pt_regs);
-	head = regs->gprs[15];
-	head_sf = (struct stack_frame*)head;
-
-	if (!head_sf->back_chain)
-		return;
-
-	head = head_sf->back_chain;
-
-	head = __show_trace(&depth, head,
-			    S390_lowcore.async_stack + frame_size - ASYNC_SIZE,
-			    S390_lowcore.async_stack + frame_size);
-
-	__show_trace(&depth, head, S390_lowcore.thread_info,
-		     S390_lowcore.thread_info + THREAD_SIZE);
-}

arch/s390/oprofile/init.c

···
 
 #include "../../../drivers/oprofile/oprof.h"
 
-extern void s390_backtrace(struct pt_regs * const regs, unsigned int depth);
-
 #include "hwsampler.h"
 #include "op_counter.h"
 
···
 		case 0x2097: case 0x2098: ops->cpu_type = "s390/z10"; break;
 		case 0x2817: case 0x2818: ops->cpu_type = "s390/z196"; break;
 		case 0x2827: case 0x2828: ops->cpu_type = "s390/zEC12"; break;
+		case 0x2964: case 0x2965: ops->cpu_type = "s390/z13"; break;
 		default: return -ENODEV;
 		}
 	}
···
 static void oprofile_hwsampler_exit(void)
 {
 	hwsampler_shutdown();
+}
+
+static int __s390_backtrace(void *data, unsigned long address)
+{
+	unsigned int *depth = data;
+
+	if (*depth == 0)
+		return 1;
+	(*depth)--;
+	oprofile_add_trace(address);
+	return 0;
+}
+
+static void s390_backtrace(struct pt_regs *regs, unsigned int depth)
+{
+	if (user_mode(regs))
+		return;
+	dump_trace(__s390_backtrace, &depth, NULL, regs->gprs[15]);
 }
 
 int __init oprofile_arch_init(struct oprofile_operations *ops)

arch/s390/pci/pci.c

···
 
 int pcibios_add_device(struct pci_dev *pdev)
 {
-	struct zpci_dev *zdev = to_zpci(pdev);
 	struct resource *res;
 	int i;
 
-	zdev->pdev = pdev;
 	pdev->dev.groups = zpci_attr_groups;
 	zpci_map_resources(pdev);
 
···
 {
 	struct zpci_dev *zdev = to_zpci(pdev);
 
-	zdev->pdev = pdev;
-	zpci_debug_init_device(zdev);
+	zpci_debug_init_device(zdev, dev_name(&pdev->dev));
 	zpci_fmb_enable_device(zdev);
 
 	return pci_enable_resources(pdev, mask);
···
 
 	zpci_fmb_disable_device(zdev);
 	zpci_debug_exit_device(zdev);
-	zdev->pdev = NULL;
 }
 
 #ifdef CONFIG_HIBERNATE_CALLBACKS
···
 
 static int zpci_mem_init(void)
 {
+	BUILD_BUG_ON(!is_power_of_2(__alignof__(struct zpci_fmb)) ||
+		     __alignof__(struct zpci_fmb) < sizeof(struct zpci_fmb));
+
 	zdev_fmb_cache = kmem_cache_create("PCI_FMB_cache", sizeof(struct zpci_fmb),
-				16, 0, NULL);
+					   __alignof__(struct zpci_fmb), 0, NULL);
 	if (!zdev_fmb_cache)
 		goto error_fmb;
 

arch/s390/pci/pci_clp.c

···
 #define KMSG_COMPONENT "zpci"
 #define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
 
+#include <linux/compat.h>
 #include <linux/kernel.h>
+#include <linux/miscdevice.h>
 #include <linux/slab.h>
 #include <linux/err.h>
 #include <linux/delay.h>
 #include <linux/pci.h>
+#include <linux/uaccess.h>
 #include <asm/pci_debug.h>
 #include <asm/pci_clp.h>
+#include <asm/compat.h>
+#include <asm/clp.h>
+#include <uapi/asm/clp.h>
 
 static inline void zpci_err_clp(unsigned int rsp, int rc)
 {
···
 }
 
 /*
- * Call Logical Processor
- * Retry logic is handled by the caller.
+ * Call Logical Processor with c=1, lps=0 and command 1
+ * to get the bit mask of installed logical processors
  */
-static inline u8 clp_instr(void *data)
+static inline int clp_get_ilp(unsigned long *ilp)
+{
+	unsigned long mask;
+	int cc = 3;
+
+	asm volatile (
+		"	.insn	rrf,0xb9a00000,%[mask],%[cmd],8,0\n"
+		"0:	ipm	%[cc]\n"
+		"	srl	%[cc],28\n"
+		"1:\n"
+		EX_TABLE(0b, 1b)
+		: [cc] "+d" (cc), [mask] "=d" (mask) : [cmd] "a" (1)
+		: "cc");
+	*ilp = mask;
+	return cc;
+}
+
+/*
+ * Call Logical Processor with c=0, the give constant lps and an lpcb request.
+ */
+static inline int clp_req(void *data, unsigned int lps)
 {
 	struct { u8 _[CLP_BLK_SIZE]; } *req = data;
 	u64 ignored;
-	u8 cc;
+	int cc = 3;
 
 	asm volatile (
-		"	.insn	rrf,0xb9a00000,%[ign],%[req],0x0,0x2\n"
-		"	ipm	%[cc]\n"
+		"	.insn	rrf,0xb9a00000,%[ign],%[req],0,%[lps]\n"
+		"0:	ipm	%[cc]\n"
 		"	srl	%[cc],28\n"
-		: [cc] "=d" (cc), [ign] "=d" (ignored), "+m" (*req)
-		: [req] "a" (req)
+		"1:\n"
+		EX_TABLE(0b, 1b)
+		: [cc] "+d" (cc), [ign] "=d" (ignored), "+m" (*req)
+		: [req] "a" (req), [lps] "i" (lps)
 		: "cc");
 	return cc;
 }
···
 	rrb->response.hdr.len = sizeof(rrb->response);
 	rrb->request.pfgid = pfgid;
 
-	rc = clp_instr(rrb);
+	rc = clp_req(rrb, CLP_LPS_PCI);
 	if (!rc && rrb->response.hdr.rsp == CLP_RC_OK)
 		clp_store_query_pci_fngrp(zdev, &rrb->response);
 	else {
···
 	rrb->response.hdr.len = sizeof(rrb->response);
 	rrb->request.fh = fh;
 
-	rc = clp_instr(rrb);
+	rc = clp_req(rrb, CLP_LPS_PCI);
 	if (!rc && rrb->response.hdr.rsp == CLP_RC_OK) {
 		rc = clp_store_query_pci_fn(zdev, &rrb->response);
 		if (rc)
···
 		rrb->request.oc = command;
 		rrb->request.ndas = nr_dma_as;
 
-		rc = clp_instr(rrb);
+		rc = clp_req(rrb, CLP_LPS_PCI);
 		if (rrb->response.hdr.rsp == CLP_RC_SETPCIFN_BUSY) {
 			retries--;
 			if (retries < 0)
···
 		rrb->request.resume_token = resume_token;
 
 		/* Get PCI function handle list */
-		rc = clp_instr(rrb);
+		rc = clp_req(rrb, CLP_LPS_PCI);
 		if (rc || rrb->response.hdr.rsp != CLP_RC_OK) {
 			zpci_err("List PCI FN:\n");
 			zpci_err_clp(rrb->response.hdr.rsp, rc);
···
 	clp_free_block(rrb);
 	return rc;
 }
+
+static int clp_base_slpc(struct clp_req *req, struct clp_req_rsp_slpc *lpcb)
+{
+	unsigned long limit = PAGE_SIZE - sizeof(lpcb->request);
+
+	if (lpcb->request.hdr.len != sizeof(lpcb->request) ||
+	    lpcb->response.hdr.len > limit)
+		return -EINVAL;
+	return clp_req(lpcb, CLP_LPS_BASE) ? -EOPNOTSUPP : 0;
+}
+
+static int clp_base_command(struct clp_req *req, struct clp_req_hdr *lpcb)
+{
+	switch (lpcb->cmd) {
+	case 0x0001: /* store logical-processor characteristics */
+		return clp_base_slpc(req, (void *) lpcb);
+	default:
+		return -EINVAL;
+	}
+}
+
+static int clp_pci_slpc(struct clp_req *req, struct clp_req_rsp_slpc *lpcb)
+{
+	unsigned long limit = PAGE_SIZE - sizeof(lpcb->request);
+
+	if (lpcb->request.hdr.len != sizeof(lpcb->request) ||
+	    lpcb->response.hdr.len > limit)
+		return -EINVAL;
+	return clp_req(lpcb, CLP_LPS_PCI) ? -EOPNOTSUPP : 0;
+}
+
+static int clp_pci_list(struct clp_req *req, struct clp_req_rsp_list_pci *lpcb)
+{
+	unsigned long limit = PAGE_SIZE - sizeof(lpcb->request);
+
+	if (lpcb->request.hdr.len != sizeof(lpcb->request) ||
+	    lpcb->response.hdr.len > limit)
+		return -EINVAL;
+	if (lpcb->request.reserved2 != 0)
+		return -EINVAL;
+	return clp_req(lpcb, CLP_LPS_PCI) ? -EOPNOTSUPP : 0;
+}
+
+static int clp_pci_query(struct clp_req *req,
+			 struct clp_req_rsp_query_pci *lpcb)
+{
+	unsigned long limit = PAGE_SIZE - sizeof(lpcb->request);
+
+	if (lpcb->request.hdr.len != sizeof(lpcb->request) ||
+	    lpcb->response.hdr.len > limit)
+		return -EINVAL;
+	if (lpcb->request.reserved2 != 0 || lpcb->request.reserved3 != 0)
+		return -EINVAL;
+	return clp_req(lpcb, CLP_LPS_PCI) ? -EOPNOTSUPP : 0;
+}
+
+static int clp_pci_query_grp(struct clp_req *req,
+			     struct clp_req_rsp_query_pci_grp *lpcb)
+{
+	unsigned long limit = PAGE_SIZE - sizeof(lpcb->request);
+
+	if (lpcb->request.hdr.len != sizeof(lpcb->request) ||
+	    lpcb->response.hdr.len > limit)
+		return -EINVAL;
+	if (lpcb->request.reserved2 != 0 || lpcb->request.reserved3 != 0 ||
+	    lpcb->request.reserved4 != 0)
+		return -EINVAL;
+	return clp_req(lpcb, CLP_LPS_PCI) ? -EOPNOTSUPP : 0;
+}
+
+static int clp_pci_command(struct clp_req *req, struct clp_req_hdr *lpcb)
+{
+	switch (lpcb->cmd) {
+	case 0x0001: /* store logical-processor characteristics */
+		return clp_pci_slpc(req, (void *) lpcb);
+	case 0x0002: /* list PCI functions */
+		return clp_pci_list(req, (void *) lpcb);
+	case 0x0003: /* query PCI function */
+		return clp_pci_query(req, (void *) lpcb);
+	case 0x0004: /* query PCI function group */
+		return clp_pci_query_grp(req, (void *) lpcb);
+	default:
+		return -EINVAL;
+	}
+}
+
+static int clp_normal_command(struct clp_req *req)
+{
+	struct clp_req_hdr *lpcb;
+	void __user *uptr;
+	int rc;
+
+	rc = -EINVAL;
+	if (req->lps != 0 && req->lps != 2)
+		goto out;
+
+	rc = -ENOMEM;
+	lpcb = clp_alloc_block(GFP_KERNEL);
+	if (!lpcb)
+		goto out;
+
+	rc = -EFAULT;
+	uptr = (void __force __user *)(unsigned long) req->data_p;
+	if (copy_from_user(lpcb, uptr, PAGE_SIZE) != 0)
+		goto out_free;
+
+	rc = -EINVAL;
+	if (lpcb->fmt != 0 || lpcb->reserved1 != 0 || lpcb->reserved2 != 0)
+		goto out_free;
+
+	switch (req->lps) {
+	case 0:
+		rc = clp_base_command(req, lpcb);
+		break;
+	case 2:
+		rc = clp_pci_command(req, lpcb);
+		break;
+	}
+	if (rc)
+		goto out_free;
+
+	rc = -EFAULT;
+	if (copy_to_user(uptr, lpcb, PAGE_SIZE) != 0)
+		goto out_free;
+
+	rc = 0;
+
+out_free:
+	clp_free_block(lpcb);
+out:
+	return rc;
+}
+
+static int clp_immediate_command(struct clp_req *req)
+{
+	void __user *uptr;
+	unsigned long ilp;
+	int exists;
+
+	if (req->cmd > 1 || clp_get_ilp(&ilp) != 0)
+		return -EINVAL;
+
+	uptr = (void __force __user *)(unsigned long) req->data_p;
+	if (req->cmd == 0) {
+		/* Command code 0: test for a specific processor */
+		exists = test_bit_inv(req->lps, &ilp);
+		return put_user(exists, (int __user *) uptr);
+	}
+	/* Command code 1: return bit mask of installed processors */
+	return put_user(ilp, (unsigned long __user *) uptr);
+}
+
+static long clp_misc_ioctl(struct file *filp, unsigned int cmd,
+			   unsigned long arg)
+{
+	struct clp_req req;
+	void __user *argp;
+
+	if (cmd != CLP_SYNC)
+		return -EINVAL;
+
+	argp = is_compat_task() ? compat_ptr(arg) : (void __user *) arg;
+	if (copy_from_user(&req, argp, sizeof(req)))
+		return -EFAULT;
+	if (req.r != 0)
+		return -EINVAL;
+	return req.c ? clp_immediate_command(&req) : clp_normal_command(&req);
+}
+
+static int clp_misc_release(struct inode *inode, struct file *filp)
+{
+	return 0;
+}
+
+static const struct file_operations clp_misc_fops = {
+	.owner = THIS_MODULE,
+	.open = nonseekable_open,
+	.release = clp_misc_release,
+	.unlocked_ioctl = clp_misc_ioctl,
+	.compat_ioctl = clp_misc_ioctl,
+	.llseek = no_llseek,
+};
+
+static struct miscdevice clp_misc_device = {
+	.minor = MISC_DYNAMIC_MINOR,
+	.name = "clp",
+	.fops = &clp_misc_fops,
+};
+
+static int __init clp_misc_init(void)
+{
+	return misc_register(&clp_misc_device);
+}
+
+device_initcall(clp_misc_init);

arch/s390/pci/pci_debug.c

···
 	.release = single_release,
 };
 
-void zpci_debug_init_device(struct zpci_dev *zdev)
+void zpci_debug_init_device(struct zpci_dev *zdev, const char *name)
 {
-	zdev->debugfs_dev = debugfs_create_dir(dev_name(&zdev->pdev->dev),
-					       debugfs_root);
+	zdev->debugfs_dev = debugfs_create_dir(name, debugfs_root);
 	if (IS_ERR(zdev->debugfs_dev))
 		zdev->debugfs_dev = NULL;
 

arch/s390/pci/pci_dma.c

···
 	dma_free_cpu_table(table);
 }
 
-static unsigned long __dma_alloc_iommu(struct zpci_dev *zdev,
+static unsigned long __dma_alloc_iommu(struct device *dev,
 				       unsigned long start, int size)
 {
+	struct zpci_dev *zdev = to_zpci(to_pci_dev(dev));
 	unsigned long boundary_size;
 
-	boundary_size = ALIGN(dma_get_seg_boundary(&zdev->pdev->dev) + 1,
+	boundary_size = ALIGN(dma_get_seg_boundary(dev) + 1,
 			      PAGE_SIZE) >> PAGE_SHIFT;
 	return iommu_area_alloc(zdev->iommu_bitmap, zdev->iommu_pages,
 				start, size, 0, boundary_size, 0);
 }
 
-static unsigned long dma_alloc_iommu(struct zpci_dev *zdev, int size)
+static unsigned long dma_alloc_iommu(struct device *dev, int size)
 {
+	struct zpci_dev *zdev = to_zpci(to_pci_dev(dev));
 	unsigned long offset, flags;
 	int wrap = 0;
 
 	spin_lock_irqsave(&zdev->iommu_bitmap_lock, flags);
-	offset = __dma_alloc_iommu(zdev, zdev->next_bit, size);
+	offset = __dma_alloc_iommu(dev, zdev->next_bit, size);
 	if (offset == -1) {
 		/* wrap-around */
-		offset = __dma_alloc_iommu(zdev, 0, size);
+		offset = __dma_alloc_iommu(dev, 0, size);
 		wrap = 1;
 	}
 
···
 	return offset;
 }
 
-static void dma_free_iommu(struct zpci_dev *zdev, unsigned long offset, int size)
+static void dma_free_iommu(struct device *dev, unsigned long offset, int size)
 {
+	struct zpci_dev *zdev = to_zpci(to_pci_dev(dev));
 	unsigned long flags;
 
 	spin_lock_irqsave(&zdev->iommu_bitmap_lock, flags);
···
 
 	/* This rounds up number of pages based on size and offset */
 	nr_pages = iommu_num_pages(pa, size, PAGE_SIZE);
-	iommu_page_index = dma_alloc_iommu(zdev, nr_pages);
+	iommu_page_index = dma_alloc_iommu(dev, nr_pages);
 	if (iommu_page_index == -1) {
 		ret = -ENOSPC;
 		goto out_err;
···
 	return dma_addr + (offset & ~PAGE_MASK);
 
 out_free:
-	dma_free_iommu(zdev, iommu_page_index, nr_pages);
+	dma_free_iommu(dev, iommu_page_index, nr_pages);
 out_err:
 	zpci_err("map error:\n");
 	zpci_err_dma(ret, pa);
···
 
 	atomic64_add(npages, &zdev->unmapped_pages);
 	iommu_page_index = (dma_addr - zdev->start_dma) >> PAGE_SHIFT;
-	dma_free_iommu(zdev, iommu_page_index, npages);
+	dma_free_iommu(dev, iommu_page_index, npages);
 }
 
 static void *s390_dma_alloc(struct device *dev, size_t size,

arch/s390/pci/pci_event.c

···
 static void __zpci_event_error(struct zpci_ccdf_err *ccdf)
 {
 	struct zpci_dev *zdev = get_zdev_by_fid(ccdf->fid);
-	struct pci_dev *pdev = zdev ? zdev->pdev : NULL;
+	struct pci_dev *pdev = NULL;
 
 	zpci_err("error CCDF:\n");
 	zpci_err_hex(ccdf, sizeof(*ccdf));
+
+	if (zdev)
+		pdev = pci_get_slot(zdev->bus, ZPCI_DEVFN);
 
 	pr_err("%s: Event 0x%x reports an error for PCI function 0x%x\n",
 	       pdev ? pci_name(pdev) : "n/a", ccdf->pec, ccdf->fid);
···
 		return;
 
 	pdev->error_state = pci_channel_io_perm_failure;
+	pci_dev_put(pdev);
 }
 
 void zpci_event_error(void *data)
···
 static void __zpci_event_availability(struct zpci_ccdf_avail *ccdf)
 {
 	struct zpci_dev *zdev = get_zdev_by_fid(ccdf->fid);
-	struct pci_dev *pdev = zdev ? zdev->pdev : NULL;
+	struct pci_dev *pdev = NULL;
 	int ret;
+
+	if (zdev)
+		pdev = pci_get_slot(zdev->bus, ZPCI_DEVFN);
 
 	pr_info("%s: Event 0x%x reconfigured PCI function 0x%x\n",
 		pdev ? pci_name(pdev) : "n/a", ccdf->pec, ccdf->fid);
···
 	default:
 		break;
 	}
+	if (pdev)
+		pci_dev_put(pdev);
 }
 
 void zpci_event_availability(void *data)

drivers/pci/hotplug/s390_pci_hpc.c

···
 static int disable_slot(struct hotplug_slot *hotplug_slot)
 {
 	struct slot *slot = hotplug_slot->private;
+	struct pci_dev *pdev;
 	int rc;
 
 	if (!zpci_fn_configured(slot->zdev->state))
 		return -EIO;
 
-	if (slot->zdev->pdev)
-		pci_stop_and_remove_bus_device_locked(slot->zdev->pdev);
+	pdev = pci_get_slot(slot->zdev->bus, ZPCI_DEVFN);
+	if (pdev) {
+		pci_stop_and_remove_bus_device_locked(pdev);
+		pci_dev_put(pdev);
+	}
 
 	rc = zpci_disable_device(slot->zdev);
 	if (rc)

drivers/s390/block/dasd_alias.c

···
  */
 int dasd_alias_make_device_known_to_lcu(struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	unsigned long flags;
 	struct alias_server *server, *newserver;
 	struct alias_lcu *lcu, *newlcu;
 	struct dasd_uid uid;
-
-	private = (struct dasd_eckd_private *) device->private;
 
 	device->discipline->get_uid(device, &uid);
 	spin_lock_irqsave(&aliastree.lock, flags);
···
  */
 void dasd_alias_disconnect_device_from_lcu(struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	unsigned long flags;
 	struct alias_lcu *lcu;
 	struct alias_server *server;
 	int was_pending;
 	struct dasd_uid uid;
 
-	private = (struct dasd_eckd_private *) device->private;
 	lcu = private->lcu;
 	/* nothing to do if already disconnected */
 	if (!lcu)
···
 			      struct dasd_device *pos)
 {
 
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	struct alias_pav_group *group;
 	struct dasd_uid uid;
-	unsigned long flags;
 
-	private = (struct dasd_eckd_private *) device->private;
-
-	/* only lock if not already locked */
-	if (device != pos)
-		spin_lock_irqsave_nested(get_ccwdev_lock(device->cdev), flags,
-					 CDEV_NESTED_SECOND);
 	private->uid.type = lcu->uac->unit[private->uid.real_unit_addr].ua_type;
 	private->uid.base_unit_addr =
 		lcu->uac->unit[private->uid.real_unit_addr].base_ua;
 	uid = private->uid;
-
-	if (device != pos)
-		spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
 
 	/* if we have no PAV anyway, we don't need to bother with PAV groups */
 	if (lcu->pav == NO_PAV) {
···
 static void _remove_device_from_lcu(struct alias_lcu *lcu,
 				    struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	struct alias_pav_group *group;
 
-	private = (struct dasd_eckd_private *) device->private;
 	list_move(&device->alias_list, &lcu->inactive_devices);
 	group = private->pavgroup;
 	if (!group)
···
 		return 1;
 
 	return 0;
+}
+
+/*
+ * This function tries to lock all devices on an lcu via trylock
+ * return NULL on success otherwise return first failed device
+ */
+static struct dasd_device *_trylock_all_devices_on_lcu(struct alias_lcu *lcu,
+						      struct dasd_device *pos)
+
+{
+	struct alias_pav_group *pavgroup;
+	struct dasd_device *device;
+
+	list_for_each_entry(device, &lcu->active_devices, alias_list) {
+		if (device == pos)
+			continue;
+		if (!spin_trylock(get_ccwdev_lock(device->cdev)))
+			return device;
+	}
+	list_for_each_entry(device, &lcu->inactive_devices, alias_list) {
+		if (device == pos)
+			continue;
+		if (!spin_trylock(get_ccwdev_lock(device->cdev)))
+			return device;
+	}
+	list_for_each_entry(pavgroup, &lcu->grouplist, group) {
+		list_for_each_entry(device, &pavgroup->baselist, alias_list) {
+			if (device == pos)
+				continue;
+			if (!spin_trylock(get_ccwdev_lock(device->cdev)))
+				return device;
+		}
+		list_for_each_entry(device, &pavgroup->aliaslist, alias_list) {
+			if (device == pos)
+				continue;
+			if (!spin_trylock(get_ccwdev_lock(device->cdev)))
+				return device;
+		}
+	}
+	return NULL;
+}
+
+/*
+ * unlock all devices except the one that is specified as pos
+ * stop if enddev is specified and reached
+ */
+static void _unlock_all_devices_on_lcu(struct alias_lcu *lcu,
+				       struct dasd_device *pos,
+				       struct dasd_device *enddev)
+
+{
+	struct alias_pav_group *pavgroup;
+	struct dasd_device *device;
+
+	list_for_each_entry(device, &lcu->active_devices, alias_list) {
+		if (device == pos)
+			continue;
+		if (device == enddev)
+			return;
+		spin_unlock(get_ccwdev_lock(device->cdev));
+	}
+	list_for_each_entry(device, &lcu->inactive_devices, alias_list) {
+		if (device == pos)
+			continue;
+		if (device == enddev)
+			return;
+		spin_unlock(get_ccwdev_lock(device->cdev));
+	}
+	list_for_each_entry(pavgroup, &lcu->grouplist, group) {
+		list_for_each_entry(device, &pavgroup->baselist, alias_list) {
+			if (device == pos)
+				continue;
+			if (device == enddev)
+				return;
+			spin_unlock(get_ccwdev_lock(device->cdev));
+		}
+		list_for_each_entry(device, &pavgroup->aliaslist, alias_list) {
+			if (device == pos)
+				continue;
+			if (device == enddev)
+				return;
+			spin_unlock(get_ccwdev_lock(device->cdev));
+		}
+	}
+}
+
+/*
+ *  this function is needed because the locking order
+ *  device lock -> lcu lock
+ *  needs to be assured when iterating over devices in an LCU
+ *
+ *  if a device is specified in pos then the device lock is already hold
+ */
+static void _trylock_and_lock_lcu_irqsave(struct alias_lcu *lcu,
+					  struct dasd_device *pos,
+					  unsigned long *flags)
+{
+	struct dasd_device *failed;
+
+	do {
+		spin_lock_irqsave(&lcu->lock, *flags);
+		failed = _trylock_all_devices_on_lcu(lcu, pos);
+		if (failed) {
+			_unlock_all_devices_on_lcu(lcu, pos, failed);
+			spin_unlock_irqrestore(&lcu->lock, *flags);
+			cpu_relax();
+		}
+	} while (failed);
+}
+
+static void _trylock_and_lock_lcu(struct alias_lcu *lcu,
+				  struct dasd_device *pos)
+{
+	struct dasd_device *failed;
+
+	do {
+		spin_lock(&lcu->lock);
+		failed = _trylock_all_devices_on_lcu(lcu, pos);
+		if (failed) {
+			_unlock_all_devices_on_lcu(lcu, pos, failed);
+			spin_unlock(&lcu->lock);
+			cpu_relax();
+		}
+	} while (failed);
 }
 
 static int read_unit_address_configuration(struct dasd_device *device,
···
 		list_for_each_entry_safe(device, tempdev, &pavgroup->baselist,
 					 alias_list) {
 			list_move(&device->alias_list, &lcu->active_devices);
-			private = (struct dasd_eckd_private *) device->private;
+			private = device->private;
 			private->pavgroup = NULL;
 		}
 		list_for_each_entry_safe(device, tempdev, &pavgroup->aliaslist,
 					 alias_list) {
 			list_move(&device->alias_list, &lcu->active_devices);
-			private = (struct dasd_eckd_private *) device->private;
+			private = device->private;
 			private->pavgroup = NULL;
 		}
 		list_del(&pavgroup->group);
···
 	if (rc)
 		return rc;
 
-	/* need to take cdev lock before lcu lock */
-	spin_lock_irqsave_nested(get_ccwdev_lock(refdev->cdev), flags,
-				 CDEV_NESTED_FIRST);
-	spin_lock(&lcu->lock);
+	_trylock_and_lock_lcu_irqsave(lcu, NULL, &flags);
 	lcu->pav = NO_PAV;
 	for (i = 0; i < MAX_DEVICES_PER_LCU; ++i) {
 		switch (lcu->uac->unit[i].ua_type) {
···
 				 alias_list) {
 		_add_device_to_lcu(lcu, device, refdev);
 	}
-	spin_unlock(&lcu->lock);
-	spin_unlock_irqrestore(get_ccwdev_lock(refdev->cdev), flags);
+	_unlock_all_devices_on_lcu(lcu, NULL, NULL);
+	spin_unlock_irqrestore(&lcu->lock, flags);
 	return 0;
 }
 
···
 
 int dasd_alias_add_device(struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	struct alias_lcu *lcu;
 	unsigned long flags;
 	int rc;
 
-	private = (struct dasd_eckd_private *) device->private;
 	lcu = private->lcu;
 	rc = 0;
-
-	/* need to take cdev lock before lcu lock */
 	spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
 	spin_lock(&lcu->lock);
 	if (!(lcu->flags & UPDATE_PENDING)) {
···
 
 int dasd_alias_update_add_device(struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
-	private = (struct dasd_eckd_private *) device->private;
+	struct dasd_eckd_private *private = device->private;
+
 	private->lcu->flags |= UPDATE_PENDING;
 	return dasd_alias_add_device(device);
 }
 
 int dasd_alias_remove_device(struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
-	struct alias_lcu *lcu;
+	struct dasd_eckd_private *private = device->private;
+	struct alias_lcu *lcu = private->lcu;
 	unsigned long flags;
 
-	private = (struct dasd_eckd_private *) device->private;
-	lcu = private->lcu;
 	/* nothing to do if already removed */
 	if (!lcu)
 		return 0;
···
 
 struct dasd_device *dasd_alias_get_start_dev(struct dasd_device *base_device)
 {
-
+	struct dasd_eckd_private *alias_priv, *private = base_device->private;
+	struct alias_pav_group *group = private->pavgroup;
+	struct alias_lcu *lcu = private->lcu;
 	struct dasd_device *alias_device;
-	struct alias_pav_group *group;
-	struct alias_lcu *lcu;
-	struct dasd_eckd_private *private, *alias_priv;
 	unsigned long flags;
 
-	private = (struct dasd_eckd_private *) base_device->private;
-	group = private->pavgroup;
-	lcu = private->lcu;
 	if (!group || !lcu)
 		return NULL;
 	if (lcu->pav == NO_PAV ||
···
 		group->next = list_first_entry(&alias_device->alias_list,
 					       struct dasd_device, alias_list);
 	spin_unlock_irqrestore(&lcu->lock, flags);
-	alias_priv = (struct dasd_eckd_private *) alias_device->private;
+	alias_priv = alias_device->private;
 	if ((alias_priv->count < private->count) && !alias_device->stopped &&
 	    !test_bit(DASD_FLAG_OFFLINE, &alias_device->flags))
 		return alias_device;
···
 	struct alias_pav_group *pavgroup;
 	struct dasd_device *device;
 	struct dasd_eckd_private *private;
-	unsigned long flags;
 
 	/* active and inactive list can contain alias as well as base devices */
 	list_for_each_entry(device, &lcu->active_devices, alias_list) {
-		private = (struct dasd_eckd_private *) device->private;
-		spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
-		if (private->uid.type != UA_BASE_DEVICE) {
-			spin_unlock_irqrestore(get_ccwdev_lock(device->cdev),
-					       flags);
+		private = device->private;
+		if (private->uid.type != UA_BASE_DEVICE)
 			continue;
-		}
-		spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
 		dasd_schedule_block_bh(device->block);
 		dasd_schedule_device_bh(device);
 	}
 	list_for_each_entry(device, &lcu->inactive_devices, alias_list) {
-		private = (struct dasd_eckd_private *) device->private;
-		spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
-		if (private->uid.type != UA_BASE_DEVICE) {
-			spin_unlock_irqrestore(get_ccwdev_lock(device->cdev),
-					       flags);
+		private = device->private;
+		if (private->uid.type != UA_BASE_DEVICE)
 			continue;
-		}
-		spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
 		dasd_schedule_block_bh(device->block);
 		dasd_schedule_device_bh(device);
 	}
···
 	spin_lock_irqsave(&lcu->lock, flags);
 	list_for_each_entry_safe(device, temp, &lcu->active_devices,
 				 alias_list) {
-		private = (struct dasd_eckd_private *) device->private;
+		private = device->private;
 		if (private->uid.type == UA_BASE_DEVICE)
 			continue;
 		list_move(&device->alias_list, &active);
···
 		if (device == list_first_entry(&active,
 					       struct dasd_device, alias_list)) {
 			list_move(&device->alias_list, &lcu->active_devices);
-			private = (struct dasd_eckd_private *) device->private;
+			private = device->private;
 			private->pavgroup = NULL;
 		}
 	}
 	spin_unlock_irqrestore(&lcu->lock, flags);
 }
 
-static void __stop_device_on_lcu(struct dasd_device *device,
-				 struct dasd_device *pos)
-{
-	/* If pos == device then device is already locked! */
-	if (pos == device) {
-		dasd_device_set_stop_bits(pos, DASD_STOPPED_SU);
-		return;
-	}
-	spin_lock(get_ccwdev_lock(pos->cdev));
-	dasd_device_set_stop_bits(pos, DASD_STOPPED_SU);
-	spin_unlock(get_ccwdev_lock(pos->cdev));
-}
-
-/*
- * This function is called in interrupt context, so the
- * cdev lock for device is already locked!
- */
-static void _stop_all_devices_on_lcu(struct alias_lcu *lcu,
-				     struct dasd_device *device)
+static void _stop_all_devices_on_lcu(struct alias_lcu *lcu)
 {
 	struct alias_pav_group *pavgroup;
-	struct dasd_device *pos;
+	struct dasd_device *device;
 
-	list_for_each_entry(pos, &lcu->active_devices, alias_list)
-		__stop_device_on_lcu(device, pos);
-	list_for_each_entry(pos, &lcu->inactive_devices, alias_list)
-		__stop_device_on_lcu(device, pos);
+	list_for_each_entry(device, &lcu->active_devices, alias_list)
+		dasd_device_set_stop_bits(device, DASD_STOPPED_SU);
+	list_for_each_entry(device, &lcu->inactive_devices, alias_list)
+		dasd_device_set_stop_bits(device, DASD_STOPPED_SU);
 	list_for_each_entry(pavgroup, &lcu->grouplist, group) {
-		list_for_each_entry(pos, &pavgroup->baselist, alias_list)
-			__stop_device_on_lcu(device, pos);
-		list_for_each_entry(pos, &pavgroup->aliaslist, alias_list)
-			__stop_device_on_lcu(device, pos);
+		list_for_each_entry(device, &pavgroup->baselist, alias_list)
+			dasd_device_set_stop_bits(device, DASD_STOPPED_SU);
+		list_for_each_entry(device, &pavgroup->aliaslist, alias_list)
+			dasd_device_set_stop_bits(device, DASD_STOPPED_SU);
 	}
 }
 
···
 {
 	struct alias_pav_group *pavgroup;
 	struct dasd_device *device;
-	unsigned long flags;
 
-	list_for_each_entry(device, &lcu->active_devices, alias_list) {
-		spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+	list_for_each_entry(device, &lcu->active_devices, alias_list)
 		dasd_device_remove_stop_bits(device, DASD_STOPPED_SU);
-		spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
-	}
-
-	list_for_each_entry(device, &lcu->inactive_devices, alias_list) {
-		spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+	list_for_each_entry(device, &lcu->inactive_devices, alias_list)
 		dasd_device_remove_stop_bits(device, DASD_STOPPED_SU);
-		spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
-	}
-
 	list_for_each_entry(pavgroup, &lcu->grouplist, group) {
-		list_for_each_entry(device, &pavgroup->baselist, alias_list) {
-			spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+		list_for_each_entry(device, &pavgroup->baselist, alias_list)
 			dasd_device_remove_stop_bits(device, DASD_STOPPED_SU);
-			spin_unlock_irqrestore(get_ccwdev_lock(device->cdev),
-					       flags);
-		}
-		list_for_each_entry(device, &pavgroup->aliaslist, alias_list) {
-			spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
+		list_for_each_entry(device, &pavgroup->aliaslist, alias_list)
 			dasd_device_remove_stop_bits(device, DASD_STOPPED_SU);
-			spin_unlock_irqrestore(get_ccwdev_lock(device->cdev),
-					       flags);
-		}
 	}
 }
 
···
 	spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
 	reset_summary_unit_check(lcu, device, suc_data->reason);
 
-	spin_lock_irqsave(&lcu->lock, flags);
+	_trylock_and_lock_lcu_irqsave(lcu, NULL, &flags);
 	_unstop_all_devices_on_lcu(lcu);
 	_restart_all_base_devices_on_lcu(lcu);
 	/* 3. read new alias configuration */
 	_schedule_lcu_update(lcu, device);
 	lcu->suc_data.device = NULL;
 	dasd_put_device(device);
+	_unlock_all_devices_on_lcu(lcu, NULL, NULL);
 	spin_unlock_irqrestore(&lcu->lock, flags);
 }
 
···
 void dasd_alias_handle_summary_unit_check(struct dasd_device *device,
 					  struct irb *irb)
 {
+	struct dasd_eckd_private *private = device->private;
 	struct alias_lcu *lcu;
 	char reason;
-	struct dasd_eckd_private *private;
 	char *sense;
-
-	private = (struct dasd_eckd_private *) device->private;
 
 	sense = dasd_get_sense(irb);
 	if (sense) {
···
 			    " unit check (no lcu structure)");
 		return;
 	}
-	spin_lock(&lcu->lock);
-	_stop_all_devices_on_lcu(lcu, device);
-	/* prepare for lcu_update */
-	private->lcu->flags |= NEED_UAC_UPDATE | UPDATE_PENDING;
+	_trylock_and_lock_lcu(lcu, device);
 	/* If this device is about to be removed just return and wait for
 	 * the next interrupt on a different device
 	 */
···
 		DBF_DEV_EVENT(DBF_WARNING, device, "%s",
 			    "device is in offline processing,"
 			    " don't do summary unit check handling");
+		_unlock_all_devices_on_lcu(lcu, device, NULL);
 		spin_unlock(&lcu->lock);
 		return;
 	}
···
 		DBF_DEV_EVENT(DBF_WARNING, device, "%s",
 			    "previous instance of summary unit check worker"
 			    " still pending");
+		_unlock_all_devices_on_lcu(lcu, device, NULL);
 		spin_unlock(&lcu->lock);
 		return ;
 	}
+	_stop_all_devices_on_lcu(lcu);
+	/* prepare for lcu_update */
+	private->lcu->flags |= NEED_UAC_UPDATE | UPDATE_PENDING;
 	lcu->suc_data.reason = reason;
 	lcu->suc_data.device = device;
 	dasd_get_device(device);
+	_unlock_all_devices_on_lcu(lcu, device, NULL);
 	spin_unlock(&lcu->lock);
 	if (!schedule_work(&lcu->suc_data.worker))
 		dasd_put_device(device);

drivers/s390/block/dasd_devmap.c

···
 		else if (len == 8 && !strncmp(str, "failfast", 8))
 			features |= DASD_FEATURE_FAILFAST;
 		else {
-			pr_warning("%*s is not a supported device option\n",
-				   len, str);
+			pr_warn("%*s is not a supported device option\n",
+				len, str);
 			rc = -EINVAL;
 		}
 		str += len;
···
 		str++;
 	}
 	if (*str != ')') {
-		pr_warning("A closing parenthesis ')' is missing in the "
-			   "dasd= parameter\n");
+		pr_warn("A closing parenthesis ')' is missing in the dasd= parameter\n");
 		rc = -EINVAL;
 	} else
 		str++;
···
 		return str + 1;
 	if (*str == '\0')
 		return str;
-	pr_warning("The dasd= parameter value %s has an invalid ending\n",
-		   str);
+	pr_warn("The dasd= parameter value %s has an invalid ending\n", str);
 	return ERR_PTR(-EINVAL);
 }
 

drivers/s390/block/dasd_diag.c

···
 mdsk_init_io(struct dasd_device *device, unsigned int blocksize,
 	     blocknum_t offset, blocknum_t *end_block)
 {
-	struct dasd_diag_private *private;
-	struct dasd_diag_init_io *iib;
+	struct dasd_diag_private *private = device->private;
+	struct dasd_diag_init_io *iib = &private->iib;
 	int rc;
 
-	private = (struct dasd_diag_private *) device->private;
-	iib = &private->iib;
 	memset(iib, 0, sizeof (struct dasd_diag_init_io));
 
 	iib->dev_nr = private->dev_id.devno;
···
 static inline int
 mdsk_term_io(struct dasd_device * device)
 {
-	struct dasd_diag_private *private;
-	struct dasd_diag_init_io *iib;
+	struct dasd_diag_private *private = device->private;
+	struct dasd_diag_init_io *iib = &private->iib;
 	int rc;
 
-	private = (struct dasd_diag_private *) device->private;
-	iib = &private->iib;
 	memset(iib, 0, sizeof (struct dasd_diag_init_io));
 	iib->dev_nr = private->dev_id.devno;
 	rc = dia250(iib, TERM_BIO);
···
 	rc = mdsk_init_io(device, device->block->bp_block, 0, NULL);
 	if (rc == 4) {
 		if (!(test_and_set_bit(DASD_FLAG_DEVICE_RO, &device->flags)))
-			pr_warning("%s: The access mode of a DIAG device "
-				   "changed to read-only\n",
-				   dev_name(&device->cdev->dev));
+			pr_warn("%s: The access mode of a DIAG device changed to read-only\n",
+				dev_name(&device->cdev->dev));
 		rc = 0;
 	}
 	if (rc)
-		pr_warning("%s: DIAG ERP failed with "
-			    "rc=%d\n", dev_name(&device->cdev->dev), rc);
+		pr_warn("%s: DIAG ERP failed with rc=%d\n",
+			dev_name(&device->cdev->dev), rc);
 }
 
 /* Start a given request at the device. Return zero on success, non-zero
···
 		cqr->status = DASD_CQR_ERROR;
 		return -EIO;
 	}
-	private = (struct dasd_diag_private *) device->private;
-	dreq = (struct dasd_diag_req *) cqr->data;
+	private = device->private;
+	dreq = cqr->data;
 
 	private->iob.dev_nr = private->dev_id.devno;
 	private->iob.key = 0;
···
 static int
 dasd_diag_check_device(struct dasd_device *device)
 {
-	struct dasd_block *block;
-	struct dasd_diag_private *private;
+	struct dasd_diag_private *private = device->private;
 	struct dasd_diag_characteristics *rdc_data;
-	struct dasd_diag_bio bio;
 	struct vtoc_cms_label *label;
-	blocknum_t end_block;
+	struct dasd_block *block;
+	struct dasd_diag_bio bio;
 	unsigned int sb, bsize;
+	blocknum_t end_block;
 	int rc;
 
-	private = (struct dasd_diag_private *) device->private;
 	if (private == NULL) {
-		private = kzalloc(sizeof(struct dasd_diag_private),GFP_KERNEL);
+		private = kzalloc(sizeof(*private), GFP_KERNEL);
 		if (private == NULL) {
 			DBF_DEV_EVENT(DBF_WARNING, device, "%s",
 				"Allocating memory for private DASD data "
···
 			return -ENOMEM;
 		}
 		ccw_device_get_id(device->cdev, &private->dev_id);
-		device->private = (void *) private;
+		device->private = private;
 	}
 	block = dasd_alloc_block();
 	if (IS_ERR(block)) {
···
 	block->base = device;
 
 	/* Read Device Characteristics */
-	rdc_data = (void *) &(private->rdc_data);
+	rdc_data = &private->rdc_data;
 	rdc_data->dev_nr = private->dev_id.devno;
 	rdc_data->rdc_len = sizeof (struct dasd_diag_characteristics);
 
···
 		private->pt_block = 2;
 		break;
 	default:
-		pr_warning("%s: Device type %d is not supported "
-			   "in DIAG mode\n", dev_name(&device->cdev->dev),
-			   private->rdc_data.vdev_class);
+		pr_warn("%s: Device type %d is not supported in DIAG mode\n",
+			dev_name(&device->cdev->dev),
+			private->rdc_data.vdev_class);
 		rc = -EOPNOTSUPP;
 		goto out;
 	}
···
 		private->iob.flaga = DASD_DIAG_FLAGA_DEFAULT;
 		rc = dia250(&private->iob, RW_BIO);
 		if (rc == 3) {
-			pr_warning("%s: A 64-bit DIAG call failed\n",
-				   dev_name(&device->cdev->dev));
+			pr_warn("%s: A 64-bit DIAG call failed\n",
+				dev_name(&device->cdev->dev));
 			rc = -EOPNOTSUPP;
 			goto out_label;
 		}
···
 			break;
 	}
 	if (bsize > PAGE_SIZE) {
-		pr_warning("%s: Accessing the DASD failed because of an "
-			   "incorrect format (rc=%d)\n",
-			   dev_name(&device->cdev->dev), rc);
+		pr_warn("%s: Accessing the DASD failed because of an incorrect format (rc=%d)\n",
+			dev_name(&device->cdev->dev), rc);
 		rc = -EIO;
 		goto out_label;
 	}
···
 		block->s2b_shift++;
 	rc = mdsk_init_io(device, block->bp_block, 0, NULL);
 	if (rc && (rc != 4)) {
-		pr_warning("%s: DIAG initialization failed with rc=%d\n",
-			   dev_name(&device->cdev->dev), rc);
+		pr_warn("%s: DIAG initialization failed with rc=%d\n",
+			dev_name(&device->cdev->dev), rc);
 		rc = -EIO;
 	} else {
 		if (rc == 4)
···
 dasd_diag_fill_info(struct dasd_device * device,
 		    struct dasd_information2_t * info)
 {
-	struct dasd_diag_private *private;
+	struct dasd_diag_private *private = device->private;
 
-	private = (struct dasd_diag_private *) device->private;
 	info->label_block = (unsigned int) private->pt_block;
 	info->FBA_layout = 1;
 	info->format = DASD_FORMAT_LDL;
-	info->characteristics_size = sizeof (struct dasd_diag_characteristics);
-	memcpy(info->characteristics,
-	       &((struct dasd_diag_private *) device->private)->rdc_data,
-	       sizeof (struct dasd_diag_characteristics));
+	info->characteristics_size = sizeof(private->rdc_data);
+	memcpy(info->characteristics, &private->rdc_data,
+	       sizeof(private->rdc_data));
 	info->confdata_size = 0;
 	return 0;
 }

drivers/s390/block/dasd_eckd.c

···
            struct DE_eckd_data *data,
            struct dasd_device  *device)
 {
-        struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	int rc;
 
-        private = (struct dasd_eckd_private *) device->private;
 	if (!private->rdc_data.facilities.XRC_supported)
 		return 0;
 
···
 define_extent(struct ccw1 *ccw, struct DE_eckd_data *data, unsigned int trk,
 	      unsigned int totrk, int cmd, struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	u32 begcyl, endcyl;
 	u16 heads, beghead, endhead;
 	int rc = 0;
-
-	private = (struct dasd_eckd_private *) device->private;
 
 	ccw->cmd_code = DASD_ECKD_CCW_DEFINE_EXTENT;
 	ccw->flags = 0;
···
 static int check_XRC_on_prefix(struct PFX_eckd_data *pfxdata,
 			       struct dasd_device  *device)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	int rc;
 
-	private = (struct dasd_eckd_private *) device->private;
 	if (!private->rdc_data.facilities.XRC_supported)
 		return 0;
 
···
 			  struct dasd_device *device, unsigned int reclen,
 			  unsigned int tlf)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	int sector;
 	int dn, d;
-
-	private = (struct dasd_eckd_private *) device->private;
 
 	memset(data, 0, sizeof(*data));
 	sector = 0;
···
 	u16 heads, beghead, endhead;
 	int rc = 0;
 
-	basepriv = (struct dasd_eckd_private *) basedev->private;
-	startpriv = (struct dasd_eckd_private *) startdev->private;
+	basepriv = basedev->private;
+	startpriv = startdev->private;
 	dedata = &pfxdata->define_extent;
 	lredata = &pfxdata->locate_record;
 
···
 	      unsigned int rec_on_trk, int no_rec, int cmd,
 	      struct dasd_device * device, int reclen)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	int sector;
 	int dn, d;
-
-	private = (struct dasd_eckd_private *) device->private;
 
 	DBF_DEV_EVENT(DBF_INFO, device,
 		  "Locate: trk %d, rec %d, no_rec %d, cmd %d, reclen %d",
···
  */
 static int dasd_eckd_generate_uid(struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	unsigned long flags;
 
-	private = (struct dasd_eckd_private *) device->private;
 	if (!private)
 		return -ENODEV;
 	if (!private->ned || !private->gneq)
···
 
 static int dasd_eckd_get_uid(struct dasd_device *device, struct dasd_uid *uid)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	unsigned long flags;
 
-	if (device->private) {
-		private = (struct dasd_eckd_private *)device->private;
+	if (private) {
 		spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
 		*uid = private->uid;
 		spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
···
 
 static void dasd_eckd_clear_conf_data(struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	int i;
 
-	private = (struct dasd_eckd_private *) device->private;
 	private->conf_data = NULL;
 	private->conf_len = 0;
 	for (i = 0; i < 8; i++) {
···
 	struct dasd_uid *uid;
 	char print_path_uid[60], print_device_uid[60];
 
-	private = (struct dasd_eckd_private *) device->private;
+	private = device->private;
 	path_data = &device->path_data;
 	opm = ccw_device_get_path_mask(device->cdev);
 	conf_data_saved = 0;
···
 
 static int verify_fcx_max_data(struct dasd_device *device, __u8 lpm)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	int mdc;
 	u32 fcx_max_data;
 
-	private = (struct dasd_eckd_private *) device->private;
 	if (private->fcx_max_data) {
 		mdc = ccw_device_get_mdc(device->cdev, lpm);
 		if ((mdc < 0)) {
···
 static int rebuild_device_uid(struct dasd_device *device,
 			      struct path_verification_work_data *data)
 {
-	struct dasd_eckd_private *private;
-	struct dasd_path *path_data;
-	__u8 lpm, opm;
-	int rc;
-
-	rc = -ENODEV;
-	private = (struct dasd_eckd_private *) device->private;
-	path_data = &device->path_data;
-	opm = device->path_data.opm;
+	struct dasd_eckd_private *private = device->private;
+	struct dasd_path *path_data = &device->path_data;
+	__u8 lpm, opm = path_data->opm;
+	int rc = -ENODEV;
 
 	for (lpm = 0x80; lpm; lpm >>= 1) {
 		if (!(lpm & opm))
···
 
 static int dasd_eckd_read_features(struct dasd_device *device)
 {
+	struct dasd_eckd_private *private = device->private;
 	struct dasd_psf_prssd_data *prssdp;
 	struct dasd_rssd_features *features;
 	struct dasd_ccw_req *cqr;
 	struct ccw1 *ccw;
 	int rc;
-	struct dasd_eckd_private *private;
 
-	private = (struct dasd_eckd_private *) device->private;
 	memset(&private->features, 0, sizeof(struct dasd_rssd_features));
 	cqr = dasd_smalloc_request(DASD_ECKD_MAGIC, 1 /* PSF */	+ 1 /* RSSD */,
 				   (sizeof(struct dasd_psf_prssd_data) +
···
 static int dasd_eckd_validate_server(struct dasd_device *device,
 				     unsigned long flags)
 {
-	int rc;
-	struct dasd_eckd_private *private;
-	int enable_pav;
+	struct dasd_eckd_private *private = device->private;
+	int enable_pav, rc;
 
-	private = (struct dasd_eckd_private *) device->private;
 	if (private->uid.type == UA_BASE_PAV_ALIAS ||
 	    private->uid.type == UA_HYPER_PAV_ALIAS)
 		return 0;
···
 
 static u32 get_fcx_max_data(struct dasd_device *device)
 {
-	int tpm, mdc;
+	struct dasd_eckd_private *private = device->private;
 	int fcx_in_css, fcx_in_gneq, fcx_in_features;
-	struct dasd_eckd_private *private;
+	int tpm, mdc;
 
 	if (dasd_nofcx)
 		return 0;
 	/* is transport mode supported? */
-	private = (struct dasd_eckd_private *) device->private;
 	fcx_in_css = css_general_characteristics.fcx;
 	fcx_in_gneq = private->gneq->reserved2[7] & 0x04;
 	fcx_in_features = private->features.feature[40] & 0x80;
···
 static int
 dasd_eckd_check_characteristics(struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	struct dasd_block *block;
 	struct dasd_uid temp_uid;
 	int rc, i;
···
 		dev_info(&device->cdev->dev,
 			 "The DASD is not operating in multipath mode\n");
 	}
-	private = (struct dasd_eckd_private *) device->private;
 	if (!private) {
 		private = kzalloc(sizeof(*private), GFP_KERNEL | GFP_DMA);
 		if (!private) {
···
 				 "failed\n");
 			return -ENOMEM;
 		}
-		device->private = (void *) private;
+		device->private = private;
 	} else {
 		memset(private, 0, sizeof(*private));
 	}
···
 
 static void dasd_eckd_uncheck_device(struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	int i;
 
-	private = (struct dasd_eckd_private *) device->private;
 	dasd_alias_disconnect_device_from_lcu(device);
 	private->ned = NULL;
 	private->sneq = NULL;
···
 static struct dasd_ccw_req *
 dasd_eckd_analysis_ccw(struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	struct eckd_count *count_data;
 	struct LO_eckd_data *LO_data;
 	struct dasd_ccw_req *cqr;
 	struct ccw1 *ccw;
 	int cplength, datasize;
 	int i;
-
-	private = (struct dasd_eckd_private *) device->private;
 
 	cplength = 8;
 	datasize = sizeof(struct DE_eckd_data) + 2*sizeof(struct LO_eckd_data);
···
 static void dasd_eckd_analysis_callback(struct dasd_ccw_req *init_cqr,
 					void *data)
 {
-	struct dasd_eckd_private *private;
-	struct dasd_device *device;
+	struct dasd_device *device = init_cqr->startdev;
+	struct dasd_eckd_private *private = device->private;
 
-	device = init_cqr->startdev;
-	private = (struct dasd_eckd_private *) device->private;
 	private->init_cqr_status = dasd_eckd_analysis_evaluation(init_cqr);
 	dasd_sfree_request(init_cqr, device);
 	dasd_kick_device(device);
···
 
 static int dasd_eckd_end_analysis(struct dasd_block *block)
 {
-	struct dasd_device *device;
-	struct dasd_eckd_private *private;
+	struct dasd_device *device = block->base;
+	struct dasd_eckd_private *private = device->private;
 	struct eckd_count *count_area;
 	unsigned int sb, blk_per_trk;
 	int status, i;
 	struct dasd_ccw_req *init_cqr;
 
-	device = block->base;
-	private = (struct dasd_eckd_private *) device->private;
 	status = private->init_cqr_status;
 	private->init_cqr_status = -1;
 	if (status == INIT_CQR_ERROR) {
···
 
 static int dasd_eckd_do_analysis(struct dasd_block *block)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = block->base->private;
 
-	private = (struct dasd_eckd_private *) block->base->private;
 	if (private->init_cqr_status < 0)
 		return dasd_eckd_start_analysis(block);
 	else
···
 static int
 dasd_eckd_fill_geometry(struct dasd_block *block, struct hd_geometry *geo)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = block->base->private;
 
-	private = (struct dasd_eckd_private *) block->base->private;
 	if (dasd_check_blocksize(block->bp_block) == 0) {
 		geo->sectors = recs_per_track(&private->rdc_data,
 					      0, block->bp_block);
···
 	if (!startdev)
 		startdev = base;
 
-	start_priv = (struct dasd_eckd_private *) startdev->private;
-	base_priv = (struct dasd_eckd_private *) base->private;
+	start_priv = startdev->private;
+	base_priv = base->private;
 
 	rpt = recs_per_track(&base_priv->rdc_data, 0, fdata->blksize);
 
···
 				 * when formatting CDL
 				 */
 				if ((intensity & 0x08) &&
-				    fdata->start_unit == 0) {
+				    address.cyl == 0 && address.head == 0) {
 					if (i < 3) {
 						ect->kl = 4;
 						ect->dl = sizes_trk0[i] - 4;
 					}
 				}
 				if ((intensity & 0x08) &&
-				    fdata->start_unit == 1) {
+				    address.cyl == 0 && address.head == 1) {
 					ect->kl = 44;
 					ect->dl = LABEL_SIZE - 44;
 				}
···
 	return fcp;
 }
 
-static int
-dasd_eckd_format_device(struct dasd_device *base,
-			struct format_data_t *fdata,
-			int enable_pav)
+/*
+ * Wrapper function to build a CCW request depending on input data
+ */
+static struct dasd_ccw_req *
+dasd_eckd_format_build_ccw_req(struct dasd_device *base,
+			       struct format_data_t *fdata, int enable_pav)
 {
-	struct dasd_ccw_req *cqr, *n;
-	struct dasd_block *block;
-	struct dasd_eckd_private *private;
-	struct list_head format_queue;
-	struct dasd_device *device;
-	int old_stop, format_step;
-	int step, rc = 0, sleep_rc;
+	return dasd_eckd_build_format(base, fdata, enable_pav);
+}
 
-	block = base->block;
-	private = (struct dasd_eckd_private *) base->private;
+/*
+ * Sanity checks on format_data
+ */
+static int dasd_eckd_format_sanity_checks(struct dasd_device *base,
+					  struct format_data_t *fdata)
+{
+	struct dasd_eckd_private *private = base->private;
 
-	/* Sanity checks. */
 	if (fdata->start_unit >=
 	    (private->real_cyl * private->rdc_data.trk_per_cyl)) {
 		dev_warn(&base->cdev->dev,
···
 			 fdata->blksize);
 		return -EINVAL;
 	}
+	return 0;
+}
+
+/*
+ * This function will process format_data originally coming from an IOCTL
+ */
+static int dasd_eckd_format_process_data(struct dasd_device *base,
+					 struct format_data_t *fdata,
+					 int enable_pav)
+{
+	struct dasd_eckd_private *private = base->private;
+	struct dasd_ccw_req *cqr, *n;
+	struct list_head format_queue;
+	struct dasd_device *device;
+	int old_start, old_stop, format_step;
+	int step, retry;
+	int rc;
+
+	rc = dasd_eckd_format_sanity_checks(base, fdata);
+	if (rc)
+		return rc;
 
 	INIT_LIST_HEAD(&format_queue);
 
+	old_start = fdata->start_unit;
 	old_stop = fdata->stop_unit;
-	while (fdata->start_unit <= 1) {
-		fdata->stop_unit = fdata->start_unit;
-		cqr = dasd_eckd_build_format(base, fdata, enable_pav);
-		list_add(&cqr->blocklist, &format_queue);
 
-		fdata->stop_unit = old_stop;
-		fdata->start_unit++;
+	format_step = DASD_CQR_MAX_CCW / recs_per_track(&private->rdc_data, 0,
+							fdata->blksize);
+	do {
+		retry = 0;
+		while (fdata->start_unit <= old_stop) {
+			step = fdata->stop_unit - fdata->start_unit + 1;
+			if (step > format_step) {
+				fdata->stop_unit =
+					fdata->start_unit + format_step - 1;
+			}
 
-		if (fdata->start_unit > fdata->stop_unit)
-			goto sleep;
-	}
+			cqr = dasd_eckd_format_build_ccw_req(base, fdata,
+							     enable_pav);
+			if (IS_ERR(cqr)) {
+				rc = PTR_ERR(cqr);
+				if (rc == -ENOMEM) {
+					if (list_empty(&format_queue))
+						goto out;
+					/*
+					 * not enough memory available, start
+					 * requests retry after first requests
+					 * were finished
+					 */
+					retry = 1;
+					break;
+				}
+				goto out_err;
+			}
+			list_add_tail(&cqr->blocklist, &format_queue);
 
-retry:
-	format_step = 255 / recs_per_track(&private->rdc_data, 0,
-					   fdata->blksize);
-	while (fdata->start_unit <= old_stop) {
-		step = fdata->stop_unit - fdata->start_unit + 1;
-		if (step > format_step)
-			fdata->stop_unit = fdata->start_unit + format_step - 1;
-
-		cqr = dasd_eckd_build_format(base, fdata, enable_pav);
-		if (IS_ERR(cqr)) {
-			if (PTR_ERR(cqr) == -ENOMEM) {
-				/*
-				 * not enough memory available
-				 * go to out and start requests
-				 * retry after first requests were finished
-				 */
-				fdata->stop_unit = old_stop;
-				goto sleep;
-			} else
-				return PTR_ERR(cqr);
+			fdata->start_unit = fdata->stop_unit + 1;
+			fdata->stop_unit = old_stop;
 		}
-		list_add(&cqr->blocklist, &format_queue);
 
-		fdata->start_unit = fdata->stop_unit + 1;
-		fdata->stop_unit = old_stop;
-	}
+		rc = dasd_sleep_on_queue(&format_queue);
 
-sleep:
-	sleep_rc = dasd_sleep_on_queue(&format_queue);
+out_err:
+		list_for_each_entry_safe(cqr, n, &format_queue, blocklist) {
+			device = cqr->startdev;
+			private = device->private;
+			if (cqr->status == DASD_CQR_FAILED)
+				rc = -EIO;
+			list_del_init(&cqr->blocklist);
+			dasd_sfree_request(cqr, device);
+			private->count--;
+		}
 
-	list_for_each_entry_safe(cqr, n, &format_queue, blocklist) {
-		device = cqr->startdev;
-		private = (struct dasd_eckd_private *) device->private;
-		if (cqr->status == DASD_CQR_FAILED)
-			rc = -EIO;
-		list_del_init(&cqr->blocklist);
-		dasd_sfree_request(cqr, device);
-		private->count--;
-	}
+		if (rc)
+			goto out;
 
-	if (sleep_rc)
-		return sleep_rc;
+	} while (retry);
 
-	/*
-	 * in case of ENOMEM we need to retry after
-	 * first requests are finished
-	 */
-	if (fdata->start_unit <= fdata->stop_unit)
-		goto retry;
+out:
+	fdata->start_unit = old_start;
+	fdata->stop_unit = old_stop;
 
 	return rc;
+}
+
+static int dasd_eckd_format_device(struct dasd_device *base,
+				   struct format_data_t *fdata, int enable_pav)
+{
+	return dasd_eckd_format_process_data(base, fdata, enable_pav);
 }
 
 static void dasd_eckd_handle_terminated_request(struct dasd_ccw_req *cqr)
···
 {
 	char mask;
 	char *sense = NULL;
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 
-	private = (struct dasd_eckd_private *) device->private;
 	/* first of all check for state change pending interrupt */
 	mask = DEV_STAT_ATTENTION | DEV_STAT_DEV_END | DEV_STAT_UNIT_EXCEP;
 	if ((scsw_dstat(&irb->scsw) & mask) == mask) {
···
 	struct dasd_device *basedev;
 
 	basedev = block->base;
-	private = (struct dasd_eckd_private *) basedev->private;
+	private = basedev->private;
 	if (rq_data_dir(req) == READ)
 		cmd = DASD_ECKD_CCW_READ_MT;
 	else if (rq_data_dir(req) == WRITE)
···
 
 
 	/* setup prefix data */
-	basepriv = (struct dasd_eckd_private *) basedev->private;
-	startpriv = (struct dasd_eckd_private *) startdev->private;
+	basepriv = basedev->private;
+	startpriv = startdev->private;
 	dedata = &pfxdata.define_extent;
 	lredata = &pfxdata.locate_record;
 
···
 	struct dasd_ccw_req *cqr;
 
 	basedev = block->base;
-	private = (struct dasd_eckd_private *) basedev->private;
+	private = basedev->private;
 
 	/* Calculate number of blocks/records per track. */
 	blksize = block->bp_block;
···
 
 	if (!dasd_page_cache)
 		goto out;
-	private = (struct dasd_eckd_private *) cqr->block->base->private;
+	private = cqr->block->base->private;
 	blksize = cqr->block->bp_block;
 	blk_per_trk = recs_per_track(&private->rdc_data, 0, blksize);
 	recid = blk_rq_pos(req) >> cqr->block->s2b_shift;
···
 	startdev = dasd_alias_get_start_dev(base);
 	if (!startdev)
 		startdev = base;
-	private = (struct dasd_eckd_private *) startdev->private;
+	private = startdev->private;
 	if (private->count >= DASD_ECKD_CHANQ_MAX_SIZE)
 		return ERR_PTR(-EBUSY);
 
···
 	unsigned long flags;
 
 	spin_lock_irqsave(get_ccwdev_lock(cqr->memdev->cdev), flags);
-	private = (struct dasd_eckd_private *) cqr->memdev->private;
+	private = cqr->memdev->private;
 	private->count--;
 	spin_unlock_irqrestore(get_ccwdev_lock(cqr->memdev->cdev), flags);
 	return dasd_eckd_free_cp(cqr, req);
···
 dasd_eckd_fill_info(struct dasd_device * device,
 		    struct dasd_information2_t * info)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 
-	private = (struct dasd_eckd_private *) device->private;
 	info->label_block = 2;
 	info->FBA_layout = private->uses_cdl ? 0 : 1;
 	info->format = private->uses_cdl ? DASD_FORMAT_CDL : DASD_FORMAT_LDL;
-	info->characteristics_size = sizeof(struct dasd_eckd_characteristics);
+	info->characteristics_size = sizeof(private->rdc_data);
 	memcpy(info->characteristics, &private->rdc_data,
-	       sizeof(struct dasd_eckd_characteristics));
+	       sizeof(private->rdc_data));
 	info->confdata_size = min((unsigned long)private->conf_len,
 				  sizeof(info->configuration_data));
 	memcpy(info->configuration_data, private->conf_data,
···
 static int
 dasd_eckd_get_attrib(struct dasd_device *device, void __user *argp)
 {
-	struct dasd_eckd_private *private =
-		(struct dasd_eckd_private *)device->private;
+	struct dasd_eckd_private *private = device->private;
 	struct attrib_data_t attrib = private->attrib;
 	int rc;
 
···
 static int
 dasd_eckd_set_attrib(struct dasd_device *device, void __user *argp)
 {
-	struct dasd_eckd_private *private =
-		(struct dasd_eckd_private *)device->private;
+	struct dasd_eckd_private *private = device->private;
 	struct attrib_data_t attrib;
 
 	if (!capable(CAP_SYS_ADMIN))
···
 
 static int dasd_eckd_restore_device(struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	struct dasd_eckd_characteristics temp_rdc_data;
 	int rc;
 	struct dasd_uid temp_uid;
 	unsigned long flags;
 	unsigned long cqr_flags = 0;
-
-	private = (struct dasd_eckd_private *) device->private;
 
 	/* Read Configuration Data */
 	rc = dasd_eckd_read_conf(device);
···
 
 static int dasd_eckd_reload_device(struct dasd_device *device)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	int rc, old_base;
 	char print_uid[60];
 	struct dasd_uid uid;
 	unsigned long flags;
-
-	private = (struct dasd_eckd_private *) device->private;
 
 	spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
 	old_base = private->uid.base_unit_addr;
···
 {
 	struct dasd_rssd_messages *message_buf;
 	struct dasd_psf_prssd_data *prssdp;
-	struct dasd_eckd_private *private;
 	struct dasd_ccw_req *cqr;
 	struct ccw1 *ccw;
 	int rc;
 
-	private = (struct dasd_eckd_private *) device->private;
 	cqr = dasd_smalloc_request(DASD_ECKD_MAGIC, 1 /* PSF */	+ 1 /* RSSD */,
 				   (sizeof(struct dasd_psf_prssd_data) +
 				    sizeof(struct dasd_rssd_messages)),
···
 						     __u8 lpum,
 						     struct dasd_cuir_message *cuir)
 {
-	struct dasd_eckd_private *private;
+	struct dasd_eckd_private *private = device->private;
 	struct dasd_conf_data *conf_data;
 	int path, pos;
 
-	private = (struct dasd_eckd_private *) device->private;
 	if (cuir->record_selector == 0)
 		goto out;
 	for (path = 0x80, pos = 0; path; path >>= 1, pos++) {
···
 static int dasd_eckd_cuir_scope(struct dasd_device *device, __u8 lpum,
 				struct dasd_cuir_message *cuir)
 {
+	struct dasd_eckd_private *private = device->private;
 	struct dasd_conf_data *ref_conf_data;
 	unsigned long bitmask = 0, mask = 0;
-	struct dasd_eckd_private *private;
 	struct dasd_conf_data *conf_data;
 	unsigned int pos, path;
 	char *ref_gneq, *gneq;
···
 	    !(cuir->neq_map[0] | cuir->neq_map[1] | cuir->neq_map[2]))
 		return lpum;
 
-	private = (struct dasd_eckd_private *) device->private;
 	/* get reference conf data */
 	ref_conf_data = dasd_eckd_get_ref_conf(device, lpum, cuir);
 	/* reference ned is determined by ned_map field */
···
 				  struct subchannel_id sch_id,
 				  struct dasd_cuir_message *cuir)
 {
+	struct dasd_eckd_private *private = device->private;
 	struct alias_pav_group *pavgroup, *tempgroup;
-	struct dasd_eckd_private *private;
 	struct dasd_device *dev, *n;
 	unsigned long paths = 0;
 	unsigned long flags;
 	int tbcpm;
 
-	private = (struct dasd_eckd_private *) device->private;
 	/* active devices */
 	list_for_each_entry_safe(dev, n, &private->lcu->active_devices,
 				 alias_list) {
···
 				 struct subchannel_id sch_id,
 				 struct dasd_cuir_message *cuir)
 {
+	struct dasd_eckd_private *private = device->private;
 	struct alias_pav_group *pavgroup, *tempgroup;
-	struct dasd_eckd_private *private;
 	struct dasd_device *dev, *n;
 	unsigned long paths = 0;
 	int tbcpm;
 
-	private = (struct dasd_eckd_private *) device->private;
 	/*
 	 * the path may have been added through a generic path event before
 	 * only trigger path verification if the path is not already in use

drivers/s390/block/dasd_fba.c

···
 static int
 dasd_fba_check_characteristics(struct dasd_device *device)
 {
-	struct dasd_block *block;
-	struct dasd_fba_private *private;
+	struct dasd_fba_private *private = device->private;
 	struct ccw_device *cdev = device->cdev;
-	int rc;
-	int readonly;
+	struct dasd_block *block;
+	int readonly, rc;
 
-	private = (struct dasd_fba_private *) device->private;
 	if (!private) {
 		private = kzalloc(sizeof(*private), GFP_KERNEL | GFP_DMA);
 		if (!private) {
···
 				 "data failed\n");
 			return -ENOMEM;
 		}
-		device->private = (void *) private;
+		device->private = private;
 	} else {
 		memset(private, 0, sizeof(*private));
 	}
···
 
 static int dasd_fba_do_analysis(struct dasd_block *block)
 {
-	struct dasd_fba_private *private;
+	struct dasd_fba_private *private = block->base->private;
 	int sb, rc;
 
-	private = (struct dasd_fba_private *) block->base->private;
 	rc = dasd_check_blocksize(private->rdc_data.blk_size);
 	if (rc) {
 		DBF_DEV_EVENT(DBF_WARNING, block->base, "unknown blocksize %d",
···
 					      struct dasd_block *block,
 					      struct request *req)
 {
-	struct dasd_fba_private *private;
+	struct dasd_fba_private *private = block->base->private;
 	unsigned long *idaws;
 	struct LO_fba_data *LO_data;
 	struct dasd_ccw_req *cqr;
···
 	unsigned int blksize, off;
 	unsigned char cmd;
 
-	private = (struct dasd_fba_private *) block->base->private;
 	if (rq_data_dir(req) == READ) {
 		cmd = DASD_FBA_CCW_READ;
 	} else if (rq_data_dir(req) == WRITE) {
···
 static int
 dasd_fba_free_cp(struct dasd_ccw_req *cqr, struct request *req)
 {
-	struct dasd_fba_private *private;
+	struct dasd_fba_private *private = cqr->block->base->private;
 	struct ccw1 *ccw;
 	struct req_iterator iter;
 	struct bio_vec bv;
···
 
 	if (!dasd_page_cache)
 		goto out;
-	private = (struct dasd_fba_private *) cqr->block->base->private;
 	blksize = cqr->block->bp_block;
 	ccw = cqr->cpaddr;
 	/* Skip over define extent & locate record. */
···
 dasd_fba_fill_info(struct dasd_device * device,
 		   struct dasd_information2_t * info)
 {
+	struct dasd_fba_private *private = device->private;
+
 	info->label_block = 1;
 	info->FBA_layout = 1;
 	info->format = DASD_FORMAT_LDL;
-	info->characteristics_size = sizeof(struct dasd_fba_characteristics);
-	memcpy(info->characteristics,
-	       &((struct dasd_fba_private *) device->private)->rdc_data,
-	       sizeof (struct dasd_fba_characteristics));
+	info->characteristics_size = sizeof(private->rdc_data);
+	memcpy(info->characteristics, &private->rdc_data,
+	       sizeof(private->rdc_data));
 	info->confdata_size = 0;
 	return 0;
 }

drivers/s390/block/dasd_genhd.c

···
 	/* Register to static dasd major 94 */
 	rc = register_blkdev(DASD_MAJOR, "dasd");
 	if (rc != 0) {
-		pr_warning("Registering the device driver with major number "
-			   "%d failed\n", DASD_MAJOR);
+		pr_warn("Registering the device driver with major number %d failed\n",
+			DASD_MAJOR);
 		return rc;
 	}
 	return 0;

drivers/s390/block/dasd_int.h

···
 typedef struct dasd_ccw_req *(*dasd_erp_fn_t) (struct dasd_ccw_req *);
 
 /*
+ * A single CQR can only contain a maximum of 255 CCWs. It is limited by
+ * the locate record and locate record extended count value which can only hold
+ * 1 Byte max.
+ */
+#define DASD_CQR_MAX_CCW 255
+
+/*
  * Unique identifier for dasd device.
  */
 #define UA_NOT_CONFIGURED  0x00
···
 	/* Device discipline stuff. */
 	struct dasd_discipline *discipline;
 	struct dasd_discipline *base_discipline;
-	char *private;
+	void *private;
 	struct dasd_path path_data;
 
 	/* Device state and target state. */

drivers/s390/block/dasd_ioctl.c

···
 dasd_format(struct dasd_block *block, struct format_data_t *fdata)
 {
 	struct dasd_device *base;
-	int enable_pav = 1;
-	int rc, retries;
-	int start, stop;
+	int rc;
 
 	base = block->base;
 	if (base->discipline->format_device == NULL)
···
 		bdput(bdev);
 	}
 
-	retries = 255;
-	/* backup start- and endtrack for retries */
-	start = fdata->start_unit;
-	stop = fdata->stop_unit;
-	do {
-		rc = base->discipline->format_device(base, fdata, enable_pav);
-		if (rc) {
-			if (rc == -EAGAIN) {
-				retries--;
-				/* disable PAV in case of errors */
-				enable_pav = 0;
-				fdata->start_unit = start;
-				fdata->stop_unit = stop;
-			} else
-				return rc;
-		} else
-			/* success */
-			break;
-	} while (retries);
+	rc = base->discipline->format_device(base, fdata, 1);
+	if (rc == -EAGAIN)
+		rc = base->discipline->format_device(base, fdata, 0);
 
-	if (!retries)
-		return -EIO;
-	else
-		return 0;
+	return rc;
 }
 
 /*
···
 		return -EFAULT;
 	}
 	if (bdev != bdev->bd_contains) {
-		pr_warning("%s: The specified DASD is a partition and cannot "
-			   "be formatted\n",
-			   dev_name(&base->cdev->dev));
+		pr_warn("%s: The specified DASD is a partition and cannot be formatted\n",
+			dev_name(&base->cdev->dev));
 		dasd_put_device(base);
 		return -EINVAL;
 	}

drivers/s390/block/dasd_proc.c

···
 	return user_len;
 out_parse_error:
 	rc = -EINVAL;
-	pr_warning("%s is not a supported value for /proc/dasd/statistics\n",
-		str);
+	pr_warn("%s is not a supported value for /proc/dasd/statistics\n", str);
 out_error:
 	vfree(buffer);
 	return rc;
 #else
-	pr_warning("/proc/dasd/statistics: is not activated in this kernel\n");
+	pr_warn("/proc/dasd/statistics: is not activated in this kernel\n");
 	return user_len;
 #endif				/* CONFIG_DASD_PROFILE */
 }

drivers/s390/block/dcssblk.c

···
 	dev_info = dcssblk_get_device_by_name(local_buf);
 	if (dev_info == NULL) {
 		up_write(&dcssblk_devices_sem);
-		pr_warning("Device %s cannot be removed because it is not a "
-			   "known device\n", local_buf);
+		pr_warn("Device %s cannot be removed because it is not a known device\n",
+			local_buf);
 		rc = -ENODEV;
 		goto out_buf;
 	}
 	if (atomic_read(&dev_info->use_count) != 0) {
 		up_write(&dcssblk_devices_sem);
-		pr_warning("Device %s cannot be removed while it is in "
-			   "use\n", local_buf);
+		pr_warn("Device %s cannot be removed while it is in use\n",
+			local_buf);
 		rc = -EBUSY;
 		goto out_buf;
 	}
···
 		case SEG_TYPE_SC:
 			/* cannot write to these segments */
 			if (bio_data_dir(bio) == WRITE) {
-				pr_warning("Writing to %s failed because it "
-					   "is a read-only device\n",
-					   dev_name(&dev_info->dev));
+				pr_warn("Writing to %s failed because it is a read-only device\n",
+					dev_name(&dev_info->dev));
 				goto fail;
 			}
 		}

drivers/s390/char/monreader.c

···
 	memcpy(&monpriv->msg_array[monpriv->write_index]->msg,
 	       msg, sizeof(*msg));
 	if (atomic_inc_return(&monpriv->msglim_count) == MON_MSGLIM) {
-		pr_warning("The read queue for monitor data is full\n");
+		pr_warn("The read queue for monitor data is full\n");
 		monpriv->msg_array[monpriv->write_index]->msglim_reached = 1;
 	}
 	monpriv->write_index = (monpriv->write_index + 1) % MON_MSGLIM;
···
 	if (monpriv->path) {
 		rc = iucv_path_sever(monpriv->path, user_data_sever);
 		if (rc)
-			pr_warning("Disconnecting the z/VM *MONITOR system "
-				   "service failed with rc=%i\n", rc);
+			pr_warn("Disconnecting the z/VM *MONITOR system service failed with rc=%i\n",
+				rc);
 		iucv_path_free(monpriv->path);
 	}
 
···
 	if (monpriv->path) {
 		rc = iucv_path_sever(monpriv->path, user_data_sever);
 		if (rc)
-			pr_warning("Disconnecting the z/VM *MONITOR system "
-				   "service failed with rc=%i\n", rc);
+			pr_warn("Disconnecting the z/VM *MONITOR system service failed with rc=%i\n",
+				rc);
 		iucv_path_free(monpriv->path);
 	}
 	atomic_set(&monpriv->iucv_severed, 0);

drivers/s390/char/sclp_cmd.c

···
 
 	/* Check response. */
 	if (request->status != SCLP_REQ_DONE) {
-		pr_warning("sync request failed (cmd=0x%08x, "
-			   "status=0x%02x)\n", cmd, request->status);
+		pr_warn("sync request failed (cmd=0x%08x, status=0x%02x)\n",
+			cmd, request->status);
 		rc = -EIO;
 	}
 out:
···
 	if (rc)
 		goto out;
 	if (sccb->header.response_code != 0x0010) {
-		pr_warning("readcpuinfo failed (response=0x%04x)\n",
-			   sccb->header.response_code);
+		pr_warn("readcpuinfo failed (response=0x%04x)\n",
+			sccb->header.response_code);
 		rc = -EIO;
 		goto out;
 	}
···
 	case 0x0120:
 		break;
 	default:
-		pr_warning("configure cpu failed (cmd=0x%08x, "
-			   "response=0x%04x)\n", cmd,
-			   sccb->header.response_code);
+		pr_warn("configure cpu failed (cmd=0x%08x, response=0x%04x)\n",
+			cmd, sccb->header.response_code);
 		rc = -EIO;
 		break;
 	}
···
 	case 0x0120:
 		break;
 	default:
-		pr_warning("assign storage failed (cmd=0x%08x, "
-			   "response=0x%04x, rn=0x%04x)\n", cmd,
-			   sccb->header.response_code, rn);
+		pr_warn("assign storage failed (cmd=0x%08x, response=0x%04x, rn=0x%04x)\n",
+			cmd, sccb->header.response_code, rn);
 		rc = -EIO;
 		break;
 	}
···
 	case 0x0450:
 		break;
 	default:
-		pr_warning("configure channel-path failed "
-			   "(cmd=0x%08x, response=0x%04x)\n", cmd,
-			   sccb->header.response_code);
+		pr_warn("configure channel-path failed (cmd=0x%08x, response=0x%04x)\n",
+			cmd, sccb->header.response_code);
 		rc = -EIO;
 		break;
 	}
···
 	if (rc)
 		goto out;
 	if (sccb->header.response_code != 0x0010) {
-		pr_warning("read channel-path info failed "
-			   "(response=0x%04x)\n", sccb->header.response_code);
+		pr_warn("read channel-path info failed (response=0x%04x)\n",
+			sccb->header.response_code);
 		rc = -EIO;
 		goto out;
 	}

drivers/s390/char/sclp_cpi_sys.c

···
 	wait_for_completion(&completion);
 
 	if (req->status != SCLP_REQ_DONE) {
-		pr_warning("request failed (status=0x%02x)\n",
-			   req->status);
+		pr_warn("request failed (status=0x%02x)\n", req->status);
 		rc = -EIO;
 		goto out_free_req;
 	}
 
 	response = ((struct cpi_sccb *) req->sccb)->header.response_code;
 	if (response != 0x0020) {
-		pr_warning("request failed with response code 0x%x\n",
-			   response);
+		pr_warn("request failed with response code 0x%x\n", response);
 		rc = -EIO;
 	}
 

drivers/s390/char/tape_core.c

···
 			 */
 			DBF_EVENT(3, "(%08x): Drive in use vanished!\n",
 				device->cdev_id);
-			pr_warning("%s: A tape unit was detached while in "
-				   "use\n", dev_name(&device->cdev->dev));
+			pr_warn("%s: A tape unit was detached while in use\n",
+				dev_name(&device->cdev->dev));
 			tape_state_set(device, TS_NOT_OPER);
 			__tape_discard_requests(device);
 			spin_unlock_irq(get_ccwdev_lock(device->cdev));

drivers/s390/char/vmlogrdr.c

···
 	if (logptr->autorecording) {
 		ret = vmlogrdr_recording(logptr,1,logptr->autopurge);
 		if (ret)
-			pr_warning("vmlogrdr: failed to start "
-				   "recording automatically\n");
+			pr_warn("vmlogrdr: failed to start recording automatically\n");
 	}
 
 	/* create connection to the system service */
···
 	if (logptr->autorecording) {
 		ret = vmlogrdr_recording(logptr,0,logptr->autopurge);
 		if (ret)
-			pr_warning("vmlogrdr: failed to stop "
-				   "recording automatically\n");
+			pr_warn("vmlogrdr: failed to stop recording automatically\n");
 	}
 	logptr->dev_in_use = 0;
 

drivers/s390/cio/blacklist.c

···
 {
 	if ((from_ssid > to_ssid) || ((from_ssid == to_ssid) && (from > to))) {
 		if (msgtrigger)
-			pr_warning("0.%x.%04x to 0.%x.%04x is not a valid "
-				   "range for cio_ignore\n", from_ssid, from,
-				   to_ssid, to);
+			pr_warn("0.%x.%04x to 0.%x.%04x is not a valid range for cio_ignore\n",
+				from_ssid, from, to_ssid, to);
 
 		return 1;
 	}
···
 	rc = 0;
 out:
 	if (rc && msgtrigger)
-		pr_warning("%s is not a valid device for the cio_ignore "
-			   "kernel parameter\n", str);
+		pr_warn("%s is not a valid device for the cio_ignore kernel parameter\n",
+			str);
 
 	return rc;
 }

drivers/s390/cio/ccwreq.c

···
 
 	for (chp = 0; chp < 8; chp++) {
 		if ((0x80 >> chp) & sch->schib.pmcw.lpum)
-			pr_warning("%s: No interrupt was received within %lus "
-				   "(CS=%02x, DS=%02x, CHPID=%x.%02x)\n",
-				   dev_name(&cdev->dev), req->timeout / HZ,
-				   scsw_cstat(&sch->schib.scsw),
-				   scsw_dstat(&sch->schib.scsw),
-				   sch->schid.cssid,
-				   sch->schib.pmcw.chpid[chp]);
+			pr_warn("%s: No interrupt was received within %lus (CS=%02x, DS=%02x, CHPID=%x.%02x)\n",
+				dev_name(&cdev->dev), req->timeout / HZ,
+				scsw_cstat(&sch->schib.scsw),
+				scsw_dstat(&sch->schib.scsw),
+				sch->schid.cssid,
+				sch->schib.pmcw.chpid[chp]);
 	}
 
 	if (!ccwreq_next_path(cdev)) {

drivers/s390/cio/cio.c

···
 
 	sch_no = cio_get_console_sch_no();
 	if (sch_no == -1) {
-		pr_warning("No CCW console was found\n");
+		pr_warn("No CCW console was found\n");
 		return ERR_PTR(-ENODEV);
 	}
 	init_subchannel_id(&schid);

drivers/s390/cio/device.c

···
 		   cdev->private->state == DEV_STATE_DISCONNECTED));
 	/* Inform the user if set offline failed. */
 	if (cdev->private->state == DEV_STATE_BOXED) {
-		pr_warning("%s: The device entered boxed state while "
-			   "being set offline\n", dev_name(&cdev->dev));
+		pr_warn("%s: The device entered boxed state while being set offline\n",
+			dev_name(&cdev->dev));
 	} else if (cdev->private->state == DEV_STATE_NOT_OPER) {
-		pr_warning("%s: The device stopped operating while "
-			   "being set offline\n", dev_name(&cdev->dev));
+		pr_warn("%s: The device stopped operating while being set offline\n",
+			dev_name(&cdev->dev));
 	}
 	/* Give up reference from ccw_device_set_online(). */
 	put_device(&cdev->dev);
···
 		spin_unlock_irq(cdev->ccwlock);
 		/* Inform the user that set online failed. */
 		if (cdev->private->state == DEV_STATE_BOXED) {
-			pr_warning("%s: Setting the device online failed "
-				   "because it is boxed\n",
-				   dev_name(&cdev->dev));
+			pr_warn("%s: Setting the device online failed because it is boxed\n",
+				dev_name(&cdev->dev));
 		} else if (cdev->private->state == DEV_STATE_NOT_OPER) {
-			pr_warning("%s: Setting the device online failed "
-				   "because it is not operational\n",
-				   dev_name(&cdev->dev));
+			pr_warn("%s: Setting the device online failed because it is not operational\n",
+				dev_name(&cdev->dev));
 		}
 		/* Give up online reference since onlining failed. */
 		put_device(&cdev->dev);
···
 
 	rc = chsc_siosl(sch->schid);
 	if (rc < 0) {
-		pr_warning("Logging for subchannel 0.%x.%04x failed with "
-			   "errno=%d\n",
-			   sch->schid.ssid, sch->schid.sch_no, rc);
+		pr_warn("Logging for subchannel 0.%x.%04x failed with errno=%d\n",
+			sch->schid.ssid, sch->schid.sch_no, rc);
 		return rc;
 	}
 	pr_notice("Logging for subchannel 0.%x.%04x was triggered\n",

drivers/s390/net/lcs.c

···
 			lcs_schedule_recovery(card);
 			break;
 		case LCS_CMD_STOPLAN:
-			pr_warning("Stoplan for %s initiated by LGW.\n",
-				   card->dev->name);
+			pr_warn("Stoplan for %s initiated by LGW\n",
+				card->dev->name);
 			if (card->dev)
 				netif_carrier_off(card->dev);
 			break;

drivers/s390/net/qeth_l3_main.c

···
 		return rc;
 	}
 #else
-	pr_warning("There is no IPv6 support for the layer 3 discipline\n");
+	pr_warn("There is no IPv6 support for the layer 3 discipline\n");
 #endif
 	return 0;
 }