1 files changed, 105 insertions, 66 deletions

diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h
index 4a29308..4105b82 100644
--- a/arch/s390/include/asm/pgtable.h
+++ b/arch/s390/include/asm/pgtable.h
@@ -57,6 +57,10 @@ extern unsigned long zero_page_mask;
 	 (((unsigned long)(vaddr)) &zero_page_mask))))
 #define __HAVE_COLOR_ZERO_PAGE
 
+/* TODO: s390 cannot support io_remap_pfn_range... */
+#define io_remap_pfn_range(vma, vaddr, pfn, size, prot) 	       \
+	remap_pfn_range(vma, vaddr, pfn, size, prot)
+
 #endif /* !__ASSEMBLY__ */
 
 /*
@@ -302,6 +306,7 @@ extern unsigned long MODULES_END;
 #define RCP_HC_BIT	0x00200000UL
 #define RCP_GR_BIT	0x00040000UL
 #define RCP_GC_BIT	0x00020000UL
+#define RCP_IN_BIT	0x00008000UL	/* IPTE notify bit */
 
 /* User dirty / referenced bit for KVM's migration feature */
 #define KVM_UR_BIT	0x00008000UL
@@ -344,6 +349,7 @@ extern unsigned long MODULES_END;
 #define _REGION3_ENTRY_CO	0x100	/* change-recording override	    */
 
 /* Bits in the segment table entry */
+#define _SEGMENT_ENTRY_ORIGIN_LARGE ~0xfffffUL /* large page address	    */
 #define _SEGMENT_ENTRY_ORIGIN	~0x7ffUL/* segment table origin		    */
 #define _SEGMENT_ENTRY_RO	0x200	/* page protection bit		    */
 #define _SEGMENT_ENTRY_INV	0x20	/* invalid segment table entry	    */
@@ -368,6 +374,7 @@ extern unsigned long MODULES_END;
 #define RCP_HC_BIT	0x0020000000000000UL
 #define RCP_GR_BIT	0x0004000000000000UL
 #define RCP_GC_BIT	0x0002000000000000UL
+#define RCP_IN_BIT	0x0000800000000000UL	/* IPTE notify bit */
 
 /* User dirty / referenced bit for KVM's migration feature */
 #define KVM_UR_BIT	0x0000800000000000UL
@@ -419,6 +426,13 @@ extern unsigned long MODULES_END;
 #define __S110	PAGE_RW
 #define __S111	PAGE_RW
 
+/*
+ * Segment entry (large page) protection definitions.
+ */
+#define SEGMENT_NONE	__pgprot(_HPAGE_TYPE_NONE)
+#define SEGMENT_RO	__pgprot(_HPAGE_TYPE_RO)
+#define SEGMENT_RW	__pgprot(_HPAGE_TYPE_RW)
+
 static inline int mm_exclusive(struct mm_struct *mm)
 {
 	return likely(mm == current->active_mm &&
@@ -734,35 +748,67 @@ struct gmap {
 
 /**
  * struct gmap_rmap - reverse mapping for segment table entries
- * @next: pointer to the next gmap_rmap structure in the list
+ * @gmap: pointer to the gmap_struct
  * @entry: pointer to a segment table entry
+ * @vmaddr: virtual address in the guest address space
  */
 struct gmap_rmap {
 	struct list_head list;
+	struct gmap *gmap;
 	unsigned long *entry;
+	unsigned long vmaddr;
 };
 
 /**
  * struct gmap_pgtable - gmap information attached to a page table
  * @vmaddr: address of the 1MB segment in the process virtual memory
- * @mapper: list of segment table entries maping a page table
+ * @mapper: list of segment table entries mapping a page table
  */
 struct gmap_pgtable {
 	unsigned long vmaddr;
 	struct list_head mapper;
 };
 
+/**
+ * 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 address);
+};
+
 struct gmap *gmap_alloc(struct mm_struct *mm);
 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 length);
+		     unsigned long to, unsigned long len);
 int gmap_unmap_segment(struct gmap *gmap, unsigned long to, unsigned long len);
+unsigned long __gmap_translate(unsigned long address, struct gmap *);
+unsigned long gmap_translate(unsigned long address, struct gmap *);
 unsigned long __gmap_fault(unsigned long address, struct gmap *);
 unsigned long gmap_fault(unsigned long address, struct gmap *);
 void gmap_discard(unsigned long from, unsigned long to, 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) & RCP_IN_BIT) {
+		pgste_val(pgste) &= ~RCP_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
@@ -907,26 +953,6 @@ static inline pte_t pte_mkspecial(pte_t pte)
 #ifdef CONFIG_HUGETLB_PAGE
 static inline pte_t pte_mkhuge(pte_t pte)
 {
-	/*
-	 * PROT_NONE needs to be remapped from the pte type to the ste type.
-	 * The HW invalid bit is also different for pte and ste. The pte
-	 * invalid bit happens to be the same as the ste _SEGMENT_ENTRY_LARGE
-	 * bit, so we don't have to clear it.
-	 */
-	if (pte_val(pte) & _PAGE_INVALID) {
-		if (pte_val(pte) & _PAGE_SWT)
-			pte_val(pte) |= _HPAGE_TYPE_NONE;
-		pte_val(pte) |= _SEGMENT_ENTRY_INV;
-	}
-	/*
-	 * Clear SW pte bits, there are no SW bits in a segment table entry.
-	 */
-	pte_val(pte) &= ~(_PAGE_SWT | _PAGE_SWX | _PAGE_SWC |
-			  _PAGE_SWR | _PAGE_SWW);
-	/*
-	 * Also set the change-override bit because we don't need dirty bit
-	 * tracking for hugetlbfs pages.
-	 */
 	pte_val(pte) |= (_SEGMENT_ENTRY_LARGE | _SEGMENT_ENTRY_CO);
 	return pte;
 }
@@ -1038,8 +1064,10 @@ static inline pte_t ptep_get_and_clear(struct mm_struct *mm,
 	pte_t pte;
 
 	mm->context.flush_mm = 1;
-	if (mm_has_pgste(mm))
+	if (mm_has_pgste(mm)) {
 		pgste = pgste_get_lock(ptep);
+		pgste = pgste_ipte_notify(mm, address, ptep, pgste);
+	}
 
 	pte = *ptep;
 	if (!mm_exclusive(mm))
@@ -1058,11 +1086,14 @@ static inline pte_t ptep_modify_prot_start(struct mm_struct *mm,
 					   unsigned long address,
 					   pte_t *ptep)
 {
+	pgste_t pgste;
 	pte_t pte;
 
 	mm->context.flush_mm = 1;
-	if (mm_has_pgste(mm))
-		pgste_get_lock(ptep);
+	if (mm_has_pgste(mm)) {
+		pgste = pgste_get_lock(ptep);
+		pgste_ipte_notify(mm, address, ptep, pgste);
+	}
 
 	pte = *ptep;
 	if (!mm_exclusive(mm))
@@ -1088,8 +1119,10 @@ static inline pte_t ptep_clear_flush(struct vm_area_struct *vma,
 	pgste_t pgste;
 	pte_t pte;
 
-	if (mm_has_pgste(vma->vm_mm))
+	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_ipte(address, ptep);
@@ -1117,8 +1150,11 @@ static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm,
 	pgste_t pgste;
 	pte_t pte;
 
-	if (mm_has_pgste(mm))
+	if (mm_has_pgste(mm)) {
 		pgste = pgste_get_lock(ptep);
+		if (!full)
+			pgste = pgste_ipte_notify(mm, address, ptep, pgste);
+	}
 
 	pte = *ptep;
 	if (!full)
@@ -1141,8 +1177,10 @@ static inline pte_t ptep_set_wrprotect(struct mm_struct *mm,
 
 	if (pte_write(pte)) {
 		mm->context.flush_mm = 1;
-		if (mm_has_pgste(mm))
+		if (mm_has_pgste(mm)) {
 			pgste = pgste_get_lock(ptep);
+			pgste = pgste_ipte_notify(mm, address, ptep, pgste);
+		}
 
 		if (!mm_exclusive(mm))
 			__ptep_ipte(address, ptep);
@@ -1166,8 +1204,10 @@ static inline int ptep_set_access_flags(struct vm_area_struct *vma,
 
 	if (pte_same(*ptep, entry))
 		return 0;
-	if (mm_has_pgste(vma->vm_mm))
+	if (mm_has_pgste(vma->vm_mm)) {
 		pgste = pgste_get_lock(ptep);
+		pgste = pgste_ipte_notify(vma->vm_mm, address, ptep, pgste);
+	}
 
 	__ptep_ipte(address, ptep);
 
@@ -1271,31 +1311,7 @@ static inline void __pmd_idte(unsigned long address, pmd_t *pmdp)
 	}
 }
 
-#ifdef CONFIG_TRANSPARENT_HUGEPAGE
-
-#define SEGMENT_NONE	__pgprot(_HPAGE_TYPE_NONE)
-#define SEGMENT_RO	__pgprot(_HPAGE_TYPE_RO)
-#define SEGMENT_RW	__pgprot(_HPAGE_TYPE_RW)
-
-#define __HAVE_ARCH_PGTABLE_DEPOSIT
-extern void pgtable_trans_huge_deposit(struct mm_struct *mm, pgtable_t pgtable);
-
-#define __HAVE_ARCH_PGTABLE_WITHDRAW
-extern pgtable_t pgtable_trans_huge_withdraw(struct mm_struct *mm);
-
-static inline int pmd_trans_splitting(pmd_t pmd)
-{
-	return pmd_val(pmd) & _SEGMENT_ENTRY_SPLIT;
-}
-
-static inline void set_pmd_at(struct mm_struct *mm, unsigned long addr,
-			      pmd_t *pmdp, pmd_t entry)
-{
-	if (!(pmd_val(entry) & _SEGMENT_ENTRY_INV) && MACHINE_HAS_EDAT1)
-		pmd_val(entry) |= _SEGMENT_ENTRY_CO;
-	*pmdp = entry;
-}
-
+#if defined(CONFIG_TRANSPARENT_HUGEPAGE) || defined(CONFIG_HUGETLB_PAGE)
 static inline unsigned long massage_pgprot_pmd(pgprot_t pgprot)
 {
 	/*
@@ -1316,10 +1332,11 @@ static inline pmd_t pmd_modify(pmd_t pmd, pgprot_t newprot)
 	return pmd;
 }
 
-static inline pmd_t pmd_mkhuge(pmd_t pmd)
+static inline pmd_t mk_pmd_phys(unsigned long physpage, pgprot_t pgprot)
 {
-	pmd_val(pmd) |= _SEGMENT_ENTRY_LARGE;
-	return pmd;
+	pmd_t __pmd;
+	pmd_val(__pmd) = physpage + massage_pgprot_pmd(pgprot);
+	return __pmd;
 }
 
 static inline pmd_t pmd_mkwrite(pmd_t pmd)
@@ -1329,6 +1346,34 @@ static inline pmd_t pmd_mkwrite(pmd_t pmd)
 		pmd_val(pmd) &= ~_SEGMENT_ENTRY_RO;
 	return pmd;
 }
+#endif /* CONFIG_TRANSPARENT_HUGEPAGE || CONFIG_HUGETLB_PAGE */
+
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+
+#define __HAVE_ARCH_PGTABLE_DEPOSIT
+extern void pgtable_trans_huge_deposit(struct mm_struct *mm, pgtable_t pgtable);
+
+#define __HAVE_ARCH_PGTABLE_WITHDRAW
+extern pgtable_t pgtable_trans_huge_withdraw(struct mm_struct *mm);
+
+static inline int pmd_trans_splitting(pmd_t pmd)
+{
+	return pmd_val(pmd) & _SEGMENT_ENTRY_SPLIT;
+}
+
+static inline void set_pmd_at(struct mm_struct *mm, unsigned long addr,
+			      pmd_t *pmdp, pmd_t entry)
+{
+	if (!(pmd_val(entry) & _SEGMENT_ENTRY_INV) && MACHINE_HAS_EDAT1)
+		pmd_val(entry) |= _SEGMENT_ENTRY_CO;
+	*pmdp = entry;
+}
+
+static inline pmd_t pmd_mkhuge(pmd_t pmd)
+{
+	pmd_val(pmd) |= _SEGMENT_ENTRY_LARGE;
+	return pmd;
+}
 
 static inline pmd_t pmd_wrprotect(pmd_t pmd)
 {
@@ -1425,13 +1470,6 @@ static inline void pmdp_set_wrprotect(struct mm_struct *mm,
 	}
 }
 
-static inline pmd_t mk_pmd_phys(unsigned long physpage, pgprot_t pgprot)
-{
-	pmd_t __pmd;
-	pmd_val(__pmd) = physpage + massage_pgprot_pmd(pgprot);
-	return __pmd;
-}
-
 #define pfn_pmd(pfn, pgprot)	mk_pmd_phys(__pa((pfn) << PAGE_SHIFT), (pgprot))
 #define mk_pmd(page, pgprot)	pfn_pmd(page_to_pfn(page), (pgprot))
 
@@ -1531,7 +1569,8 @@ extern int s390_enable_sie(void);
 /*
  * No page table caches to initialise
  */
-#define pgtable_cache_init()	do { } while (0)
+static inline void pgtable_cache_init(void) { }
+static inline void check_pgt_cache(void) { }
 
 #include <asm-generic/pgtable.h>