From 1aea9b3f921003f0880f0676ae85d87c9f1cb4a2 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Mon, 24 Apr 2017 18:19:10 +0200
Subject: s390/mm: implement 5 level pages tables

Add the logic to upgrade the page table for a 64-bit process to
five levels. This increases the TASK_SIZE from 8PB to 16EB-4K.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/Kconfig                 |  2 +-
 arch/s390/include/asm/page.h      |  3 ++
 arch/s390/include/asm/pgalloc.h   | 25 ++++++++++--
 arch/s390/include/asm/pgtable.h   | 80 ++++++++++++++++++++++++++++++++-------
 arch/s390/include/asm/processor.h |  4 +-
 arch/s390/include/asm/tlb.h       | 15 ++++++++
 arch/s390/mm/dump_pagetables.c    | 23 +++++++++--
 arch/s390/mm/gmap.c               |  5 ++-
 arch/s390/mm/gup.c                | 33 +++++++++++++---
 arch/s390/mm/hugetlbpage.c        | 30 +++++++++------
 arch/s390/mm/mmap.c               |  4 +-
 arch/s390/mm/pageattr.c           | 30 +++++++++++++--
 arch/s390/mm/pgalloc.c            | 57 ++++++++++++++++++----------
 arch/s390/mm/pgtable.c            |  6 ++-
 arch/s390/mm/vmem.c               | 44 ++++++++++++++++++---
 15 files changed, 289 insertions(+), 72 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
index e161fafb495b..e7ff58150e8f 100644
--- a/arch/s390/Kconfig
+++ b/arch/s390/Kconfig
@@ -184,7 +184,7 @@ config SCHED_OMIT_FRAME_POINTER
 
 config PGTABLE_LEVELS
 	int
-	default 4
+	default 5
 
 source "init/Kconfig"
 
diff --git a/arch/s390/include/asm/page.h b/arch/s390/include/asm/page.h
index 69b8a41fca84..624deaa44230 100644
--- a/arch/s390/include/asm/page.h
+++ b/arch/s390/include/asm/page.h
@@ -74,6 +74,7 @@ typedef struct { unsigned long pgste; } pgste_t;
 typedef struct { unsigned long pte; } pte_t;
 typedef struct { unsigned long pmd; } pmd_t;
 typedef struct { unsigned long pud; } pud_t;
+typedef struct { unsigned long p4d; } p4d_t;
 typedef struct { unsigned long pgd; } pgd_t;
 typedef pte_t *pgtable_t;
 
@@ -82,12 +83,14 @@ typedef pte_t *pgtable_t;
 #define pte_val(x)	((x).pte)
 #define pmd_val(x)	((x).pmd)
 #define pud_val(x)	((x).pud)
+#define p4d_val(x)	((x).p4d)
 #define pgd_val(x)      ((x).pgd)
 
 #define __pgste(x)	((pgste_t) { (x) } )
 #define __pte(x)        ((pte_t) { (x) } )
 #define __pmd(x)        ((pmd_t) { (x) } )
 #define __pud(x)	((pud_t) { (x) } )
+#define __p4d(x)	((p4d_t) { (x) } )
 #define __pgd(x)        ((pgd_t) { (x) } )
 #define __pgprot(x)     ((pgprot_t) { (x) } )
 
diff --git a/arch/s390/include/asm/pgalloc.h b/arch/s390/include/asm/pgalloc.h
index 166f703dad7c..bb0ff1bb0c4a 100644
--- a/arch/s390/include/asm/pgalloc.h
+++ b/arch/s390/include/asm/pgalloc.h
@@ -51,12 +51,24 @@ static inline unsigned long pgd_entry_type(struct mm_struct *mm)
 		return _SEGMENT_ENTRY_EMPTY;
 	if (mm->context.asce_limit <= (1UL << 42))
 		return _REGION3_ENTRY_EMPTY;
-	return _REGION2_ENTRY_EMPTY;
+	if (mm->context.asce_limit <= (1UL << 53))
+		return _REGION2_ENTRY_EMPTY;
+	return _REGION1_ENTRY_EMPTY;
 }
 
-int crst_table_upgrade(struct mm_struct *);
+int crst_table_upgrade(struct mm_struct *mm, unsigned long limit);
 void crst_table_downgrade(struct mm_struct *);
 
+static inline p4d_t *p4d_alloc_one(struct mm_struct *mm, unsigned long address)
+{
+	unsigned long *table = crst_table_alloc(mm);
+
+	if (table)
+		crst_table_init(table, _REGION2_ENTRY_EMPTY);
+	return (p4d_t *) table;
+}
+#define p4d_free(mm, p4d) crst_table_free(mm, (unsigned long *) p4d)
+
 static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long address)
 {
 	unsigned long *table = crst_table_alloc(mm);
@@ -86,9 +98,14 @@ static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd)
 	crst_table_free(mm, (unsigned long *) pmd);
 }
 
-static inline void pgd_populate(struct mm_struct *mm, pgd_t *pgd, pud_t *pud)
+static inline void pgd_populate(struct mm_struct *mm, pgd_t *pgd, p4d_t *p4d)
+{
+	pgd_val(*pgd) = _REGION1_ENTRY | __pa(p4d);
+}
+
+static inline void p4d_populate(struct mm_struct *mm, p4d_t *p4d, pud_t *pud)
 {
-	pgd_val(*pgd) = _REGION2_ENTRY | __pa(pud);
+	p4d_val(*p4d) = _REGION2_ENTRY | __pa(pud);
 }
 
 static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd)
diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h
index e6e3b887bee3..3effb26f0e1a 100644
--- a/arch/s390/include/asm/pgtable.h
+++ b/arch/s390/include/asm/pgtable.h
@@ -24,7 +24,6 @@
  * the S390 page table tree.
  */
 #ifndef __ASSEMBLY__
-#include <asm-generic/5level-fixup.h>
 #include <linux/sched.h>
 #include <linux/mm_types.h>
 #include <linux/page-flags.h>
@@ -87,12 +86,15 @@ extern unsigned long zero_page_mask;
  */
 #define PMD_SHIFT	20
 #define PUD_SHIFT	31
-#define PGDIR_SHIFT	42
+#define P4D_SHIFT	42
+#define PGDIR_SHIFT	53
 
 #define PMD_SIZE        (1UL << PMD_SHIFT)
 #define PMD_MASK        (~(PMD_SIZE-1))
 #define PUD_SIZE	(1UL << PUD_SHIFT)
 #define PUD_MASK	(~(PUD_SIZE-1))
+#define P4D_SIZE	(1UL << P4D_SHIFT)
+#define P4D_MASK	(~(P4D_SIZE-1))
 #define PGDIR_SIZE	(1UL << PGDIR_SHIFT)
 #define PGDIR_MASK	(~(PGDIR_SIZE-1))
 
@@ -105,6 +107,7 @@ extern unsigned long zero_page_mask;
 #define PTRS_PER_PTE	256
 #define PTRS_PER_PMD	2048
 #define PTRS_PER_PUD	2048
+#define PTRS_PER_P4D	2048
 #define PTRS_PER_PGD	2048
 
 #define FIRST_USER_ADDRESS  0UL
@@ -115,6 +118,8 @@ extern unsigned long zero_page_mask;
 	printk("%s:%d: bad pmd %p.\n", __FILE__, __LINE__, (void *) pmd_val(e))
 #define pud_ERROR(e) \
 	printk("%s:%d: bad pud %p.\n", __FILE__, __LINE__, (void *) pud_val(e))
+#define p4d_ERROR(e) \
+	printk("%s:%d: bad p4d %p.\n", __FILE__, __LINE__, (void *) p4d_val(e))
 #define pgd_ERROR(e) \
 	printk("%s:%d: bad pgd %p.\n", __FILE__, __LINE__, (void *) pgd_val(e))
 
@@ -310,8 +315,8 @@ static inline int is_module_addr(void *addr)
 #define _REGION3_ENTRY_SOFT_DIRTY 0x0000 /* SW region soft dirty bit */
 #endif
 
-#define _REGION_ENTRY_BITS	 0xfffffffffffff227UL
-#define _REGION_ENTRY_BITS_LARGE 0xffffffff8000fe27UL
+#define _REGION_ENTRY_BITS	 0xfffffffffffff22fUL
+#define _REGION_ENTRY_BITS_LARGE 0xffffffff8000fe2fUL
 
 /* Bits in the segment table entry */
 #define _SEGMENT_ENTRY_BITS	0xfffffffffffffe33UL
@@ -564,14 +569,14 @@ static inline void crdte(unsigned long old, unsigned long new,
  */
 static inline int pgd_present(pgd_t pgd)
 {
-	if ((pgd_val(pgd) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R2)
+	if ((pgd_val(pgd) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R1)
 		return 1;
 	return (pgd_val(pgd) & _REGION_ENTRY_ORIGIN) != 0UL;
 }
 
 static inline int pgd_none(pgd_t pgd)
 {
-	if ((pgd_val(pgd) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R2)
+	if ((pgd_val(pgd) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R1)
 		return 0;
 	return (pgd_val(pgd) & _REGION_ENTRY_INVALID) != 0UL;
 }
@@ -589,6 +594,28 @@ static inline int pgd_bad(pgd_t pgd)
 	return (pgd_val(pgd) & mask) != 0;
 }
 
+static inline int p4d_present(p4d_t p4d)
+{
+	if ((p4d_val(p4d) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R2)
+		return 1;
+	return (p4d_val(p4d) & _REGION_ENTRY_ORIGIN) != 0UL;
+}
+
+static inline int p4d_none(p4d_t p4d)
+{
+	if ((p4d_val(p4d) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R2)
+		return 0;
+	return p4d_val(p4d) == _REGION2_ENTRY_EMPTY;
+}
+
+static inline unsigned long p4d_pfn(p4d_t p4d)
+{
+	unsigned long origin_mask;
+
+	origin_mask = _REGION_ENTRY_ORIGIN;
+	return (p4d_val(p4d) & origin_mask) >> PAGE_SHIFT;
+}
+
 static inline int pud_present(pud_t pud)
 {
 	if ((pud_val(pud) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R3)
@@ -641,6 +668,13 @@ static inline int pud_bad(pud_t pud)
 	return (pud_val(pud) & ~_REGION_ENTRY_BITS) != 0;
 }
 
+static inline int p4d_bad(p4d_t p4d)
+{
+	if ((p4d_val(p4d) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R2)
+		return pud_bad(__pud(p4d_val(p4d)));
+	return (p4d_val(p4d) & ~_REGION_ENTRY_BITS) != 0;
+}
+
 static inline int pmd_present(pmd_t pmd)
 {
 	return pmd_val(pmd) != _SEGMENT_ENTRY_EMPTY;
@@ -794,8 +828,14 @@ static inline int pte_unused(pte_t pte)
 
 static inline void pgd_clear(pgd_t *pgd)
 {
-	if ((pgd_val(*pgd) & _REGION_ENTRY_TYPE_MASK) == _REGION_ENTRY_TYPE_R2)
-		pgd_val(*pgd) = _REGION2_ENTRY_EMPTY;
+	if ((pgd_val(*pgd) & _REGION_ENTRY_TYPE_MASK) == _REGION_ENTRY_TYPE_R1)
+		pgd_val(*pgd) = _REGION1_ENTRY_EMPTY;
+}
+
+static inline void p4d_clear(p4d_t *p4d)
+{
+	if ((p4d_val(*p4d) & _REGION_ENTRY_TYPE_MASK) == _REGION_ENTRY_TYPE_R2)
+		p4d_val(*p4d) = _REGION2_ENTRY_EMPTY;
 }
 
 static inline void pud_clear(pud_t *pud)
@@ -1089,6 +1129,7 @@ static inline pte_t mk_pte(struct page *page, pgprot_t pgprot)
 }
 
 #define pgd_index(address) (((address) >> PGDIR_SHIFT) & (PTRS_PER_PGD-1))
+#define p4d_index(address) (((address) >> P4D_SHIFT) & (PTRS_PER_P4D-1))
 #define pud_index(address) (((address) >> PUD_SHIFT) & (PTRS_PER_PUD-1))
 #define pmd_index(address) (((address) >> PMD_SHIFT) & (PTRS_PER_PMD-1))
 #define pte_index(address) (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE-1))
@@ -1098,19 +1139,31 @@ static inline pte_t mk_pte(struct page *page, pgprot_t pgprot)
 
 #define pmd_deref(pmd) (pmd_val(pmd) & _SEGMENT_ENTRY_ORIGIN)
 #define pud_deref(pud) (pud_val(pud) & _REGION_ENTRY_ORIGIN)
+#define p4d_deref(pud) (p4d_val(pud) & _REGION_ENTRY_ORIGIN)
 #define pgd_deref(pgd) (pgd_val(pgd) & _REGION_ENTRY_ORIGIN)
 
-static inline pud_t *pud_offset(pgd_t *pgd, unsigned long address)
+static inline p4d_t *p4d_offset(pgd_t *pgd, unsigned long address)
 {
-	pud_t *pud = (pud_t *) pgd;
-	if ((pgd_val(*pgd) & _REGION_ENTRY_TYPE_MASK) == _REGION_ENTRY_TYPE_R2)
-		pud = (pud_t *) pgd_deref(*pgd);
-	return pud  + pud_index(address);
+	p4d_t *p4d = (p4d_t *) pgd;
+
+	if ((pgd_val(*pgd) & _REGION_ENTRY_TYPE_MASK) == _REGION_ENTRY_TYPE_R1)
+		p4d = (p4d_t *) pgd_deref(*pgd);
+	return p4d + p4d_index(address);
+}
+
+static inline pud_t *pud_offset(p4d_t *p4d, unsigned long address)
+{
+	pud_t *pud = (pud_t *) p4d;
+
+	if ((p4d_val(*p4d) & _REGION_ENTRY_TYPE_MASK) == _REGION_ENTRY_TYPE_R2)
+		pud = (pud_t *) p4d_deref(*p4d);
+	return pud + pud_index(address);
 }
 
 static inline pmd_t *pmd_offset(pud_t *pud, unsigned long address)
 {
 	pmd_t *pmd = (pmd_t *) pud;
+
 	if ((pud_val(*pud) & _REGION_ENTRY_TYPE_MASK) == _REGION_ENTRY_TYPE_R3)
 		pmd = (pmd_t *) pud_deref(*pud);
 	return pmd + pmd_index(address);
@@ -1122,6 +1175,7 @@ static inline pmd_t *pmd_offset(pud_t *pud, unsigned long address)
 
 #define pmd_page(pmd) pfn_to_page(pmd_pfn(pmd))
 #define pud_page(pud) pfn_to_page(pud_pfn(pud))
+#define p4d_page(pud) pfn_to_page(p4d_pfn(p4d))
 
 /* Find an entry in the lowest level page table.. */
 #define pte_offset(pmd, addr) ((pte_t *) pmd_deref(*(pmd)) + pte_index(addr))
diff --git a/arch/s390/include/asm/processor.h b/arch/s390/include/asm/processor.h
index 60d395fdc864..f57c017a5c03 100644
--- a/arch/s390/include/asm/processor.h
+++ b/arch/s390/include/asm/processor.h
@@ -92,11 +92,11 @@ extern void execve_tail(void);
  */
 
 #define TASK_SIZE_OF(tsk)	(test_tsk_thread_flag(tsk, TIF_31BIT) ? \
-					(1UL << 31) : (1UL << 53))
+					(1UL << 31) : -PAGE_SIZE)
 #define TASK_UNMAPPED_BASE	(test_thread_flag(TIF_31BIT) ? \
 					(1UL << 30) : (1UL << 41))
 #define TASK_SIZE		TASK_SIZE_OF(current)
-#define TASK_SIZE_MAX		(1UL << 53)
+#define TASK_SIZE_MAX		(-PAGE_SIZE)
 
 #define STACK_TOP		(test_thread_flag(TIF_31BIT) ? \
 					(1UL << 31) : (1UL << 42))
diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h
index 853b2a3d8dee..7317b3108a88 100644
--- a/arch/s390/include/asm/tlb.h
+++ b/arch/s390/include/asm/tlb.h
@@ -136,6 +136,21 @@ static inline void pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd,
 	tlb_remove_table(tlb, pmd);
 }
 
+/*
+ * p4d_free_tlb frees a pud table and clears the CRSTE for the
+ * region second table entry from the tlb.
+ * If the mm uses a four level page table the single p4d is freed
+ * as the pgd. p4d_free_tlb checks the asce_limit against 8PB
+ * to avoid the double free of the p4d in this case.
+ */
+static inline void p4d_free_tlb(struct mmu_gather *tlb, p4d_t *p4d,
+				unsigned long address)
+{
+	if (tlb->mm->context.asce_limit <= (1UL << 53))
+		return;
+	tlb_remove_table(tlb, p4d);
+}
+
 /*
  * pud_free_tlb frees a pud table and clears the CRSTE for the
  * region third table entry from the tlb.
diff --git a/arch/s390/mm/dump_pagetables.c b/arch/s390/mm/dump_pagetables.c
index 1b553d847140..049c3c455b32 100644
--- a/arch/s390/mm/dump_pagetables.c
+++ b/arch/s390/mm/dump_pagetables.c
@@ -149,7 +149,7 @@ static void walk_pmd_level(struct seq_file *m, struct pg_state *st,
 }
 
 static void walk_pud_level(struct seq_file *m, struct pg_state *st,
-			   pgd_t *pgd, unsigned long addr)
+			   p4d_t *p4d, unsigned long addr)
 {
 	unsigned int prot;
 	pud_t *pud;
@@ -157,7 +157,7 @@ static void walk_pud_level(struct seq_file *m, struct pg_state *st,
 
 	for (i = 0; i < PTRS_PER_PUD && addr < max_addr; i++) {
 		st->current_address = addr;
-		pud = pud_offset(pgd, addr);
+		pud = pud_offset(p4d, addr);
 		if (!pud_none(*pud))
 			if (pud_large(*pud)) {
 				prot = pud_val(*pud) &
@@ -172,6 +172,23 @@ static void walk_pud_level(struct seq_file *m, struct pg_state *st,
 	}
 }
 
+static void walk_p4d_level(struct seq_file *m, struct pg_state *st,
+			   pgd_t *pgd, unsigned long addr)
+{
+	p4d_t *p4d;
+	int i;
+
+	for (i = 0; i < PTRS_PER_P4D && addr < max_addr; i++) {
+		st->current_address = addr;
+		p4d = p4d_offset(pgd, addr);
+		if (!p4d_none(*p4d))
+			walk_pud_level(m, st, p4d, addr);
+		else
+			note_page(m, st, _PAGE_INVALID, 2);
+		addr += P4D_SIZE;
+	}
+}
+
 static void walk_pgd_level(struct seq_file *m)
 {
 	unsigned long addr = 0;
@@ -184,7 +201,7 @@ static void walk_pgd_level(struct seq_file *m)
 		st.current_address = addr;
 		pgd = pgd_offset_k(addr);
 		if (!pgd_none(*pgd))
-			walk_pud_level(m, &st, pgd, addr);
+			walk_p4d_level(m, &st, pgd, addr);
 		else
 			note_page(m, &st, _PAGE_INVALID, 1);
 		addr += PGDIR_SIZE;
diff --git a/arch/s390/mm/gmap.c b/arch/s390/mm/gmap.c
index 7f6db1e6c048..fbd664e48098 100644
--- a/arch/s390/mm/gmap.c
+++ b/arch/s390/mm/gmap.c
@@ -537,6 +537,7 @@ int __gmap_link(struct gmap *gmap, unsigned long gaddr, unsigned long vmaddr)
 	unsigned long *table;
 	spinlock_t *ptl;
 	pgd_t *pgd;
+	p4d_t *p4d;
 	pud_t *pud;
 	pmd_t *pmd;
 	int rc;
@@ -573,7 +574,9 @@ int __gmap_link(struct gmap *gmap, unsigned long gaddr, unsigned long vmaddr)
 	mm = gmap->mm;
 	pgd = pgd_offset(mm, vmaddr);
 	VM_BUG_ON(pgd_none(*pgd));
-	pud = pud_offset(pgd, vmaddr);
+	p4d = p4d_offset(pgd, vmaddr);
+	VM_BUG_ON(p4d_none(*p4d));
+	pud = pud_offset(p4d, vmaddr);
 	VM_BUG_ON(pud_none(*pud));
 	/* large puds cannot yet be handled */
 	if (pud_large(*pud))
diff --git a/arch/s390/mm/gup.c b/arch/s390/mm/gup.c
index b7b779c40a5b..8ecc25e760fa 100644
--- a/arch/s390/mm/gup.c
+++ b/arch/s390/mm/gup.c
@@ -166,15 +166,15 @@ static int gup_huge_pud(pud_t *pudp, pud_t pud, unsigned long addr,
 	return 1;
 }
 
-static inline int gup_pud_range(pgd_t *pgdp, pgd_t pgd, unsigned long addr,
+static inline int gup_pud_range(p4d_t *p4dp, p4d_t p4d, unsigned long addr,
 		unsigned long end, int write, struct page **pages, int *nr)
 {
 	unsigned long next;
 	pud_t *pudp, pud;
 
-	pudp = (pud_t *) pgdp;
-	if ((pgd_val(pgd) & _REGION_ENTRY_TYPE_MASK) == _REGION_ENTRY_TYPE_R2)
-		pudp = (pud_t *) pgd_deref(pgd);
+	pudp = (pud_t *) p4dp;
+	if ((p4d_val(p4d) & _REGION_ENTRY_TYPE_MASK) == _REGION_ENTRY_TYPE_R2)
+		pudp = (pud_t *) p4d_deref(p4d);
 	pudp += pud_index(addr);
 	do {
 		pud = *pudp;
@@ -194,6 +194,29 @@ static inline int gup_pud_range(pgd_t *pgdp, pgd_t pgd, unsigned long addr,
 	return 1;
 }
 
+static inline int gup_p4d_range(pgd_t *pgdp, pgd_t pgd, unsigned long addr,
+		unsigned long end, int write, struct page **pages, int *nr)
+{
+	unsigned long next;
+	p4d_t *p4dp, p4d;
+
+	p4dp = (p4d_t *) pgdp;
+	if ((pgd_val(pgd) & _REGION_ENTRY_TYPE_MASK) == _REGION_ENTRY_TYPE_R1)
+		p4dp = (p4d_t *) pgd_deref(pgd);
+	p4dp += p4d_index(addr);
+	do {
+		p4d = *p4dp;
+		barrier();
+		next = p4d_addr_end(addr, end);
+		if (p4d_none(p4d))
+			return 0;
+		if (!gup_pud_range(p4dp, p4d, addr, next, write, pages, nr))
+			return 0;
+	} while (p4dp++, addr = next, addr != end);
+
+	return 1;
+}
+
 /*
  * Like get_user_pages_fast() except its IRQ-safe in that it won't fall
  * back to the regular GUP.
@@ -228,7 +251,7 @@ int __get_user_pages_fast(unsigned long start, int nr_pages, int write,
 		next = pgd_addr_end(addr, end);
 		if (pgd_none(pgd))
 			break;
-		if (!gup_pud_range(pgdp, pgd, addr, next, write, pages, &nr))
+		if (!gup_p4d_range(pgdp, pgd, addr, next, write, pages, &nr))
 			break;
 	} while (pgdp++, addr = next, addr != end);
 	local_irq_restore(flags);
diff --git a/arch/s390/mm/hugetlbpage.c b/arch/s390/mm/hugetlbpage.c
index 9b4050caa4e9..d3a5e39756f6 100644
--- a/arch/s390/mm/hugetlbpage.c
+++ b/arch/s390/mm/hugetlbpage.c
@@ -162,16 +162,20 @@ pte_t *huge_pte_alloc(struct mm_struct *mm,
 			unsigned long addr, unsigned long sz)
 {
 	pgd_t *pgdp;
+	p4d_t *p4dp;
 	pud_t *pudp;
 	pmd_t *pmdp = NULL;
 
 	pgdp = pgd_offset(mm, addr);
-	pudp = pud_alloc(mm, pgdp, addr);
-	if (pudp) {
-		if (sz == PUD_SIZE)
-			return (pte_t *) pudp;
-		else if (sz == PMD_SIZE)
-			pmdp = pmd_alloc(mm, pudp, addr);
+	p4dp = p4d_alloc(mm, pgdp, addr);
+	if (p4dp) {
+		pudp = pud_alloc(mm, p4dp, addr);
+		if (pudp) {
+			if (sz == PUD_SIZE)
+				return (pte_t *) pudp;
+			else if (sz == PMD_SIZE)
+				pmdp = pmd_alloc(mm, pudp, addr);
+		}
 	}
 	return (pte_t *) pmdp;
 }
@@ -179,16 +183,20 @@ pte_t *huge_pte_alloc(struct mm_struct *mm,
 pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr)
 {
 	pgd_t *pgdp;
+	p4d_t *p4dp;
 	pud_t *pudp;
 	pmd_t *pmdp = NULL;
 
 	pgdp = pgd_offset(mm, addr);
 	if (pgd_present(*pgdp)) {
-		pudp = pud_offset(pgdp, addr);
-		if (pud_present(*pudp)) {
-			if (pud_large(*pudp))
-				return (pte_t *) pudp;
-			pmdp = pmd_offset(pudp, addr);
+		p4dp = p4d_offset(pgdp, addr);
+		if (p4d_present(*p4dp)) {
+			pudp = pud_offset(p4dp, addr);
+			if (pud_present(*pudp)) {
+				if (pud_large(*pudp))
+					return (pte_t *) pudp;
+				pmdp = pmd_offset(pudp, addr);
+			}
 		}
 	}
 	return (pte_t *) pmdp;
diff --git a/arch/s390/mm/mmap.c b/arch/s390/mm/mmap.c
index b017daed6887..8c5f284044ef 100644
--- a/arch/s390/mm/mmap.c
+++ b/arch/s390/mm/mmap.c
@@ -120,7 +120,7 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr,
 
 check_asce_limit:
 	if (addr + len > current->mm->context.asce_limit) {
-		rc = crst_table_upgrade(mm);
+		rc = crst_table_upgrade(mm, addr + len);
 		if (rc)
 			return (unsigned long) rc;
 	}
@@ -184,7 +184,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0,
 
 check_asce_limit:
 	if (addr + len > current->mm->context.asce_limit) {
-		rc = crst_table_upgrade(mm);
+		rc = crst_table_upgrade(mm, addr + len);
 		if (rc)
 			return (unsigned long) rc;
 	}
diff --git a/arch/s390/mm/pageattr.c b/arch/s390/mm/pageattr.c
index 49e721f3645e..180481589246 100644
--- a/arch/s390/mm/pageattr.c
+++ b/arch/s390/mm/pageattr.c
@@ -229,14 +229,14 @@ static void modify_pud_page(pud_t *pudp, unsigned long addr,
 	pgt_set((unsigned long *)pudp, pud_val(new), addr, CRDTE_DTT_REGION3);
 }
 
-static int walk_pud_level(pgd_t *pgd, unsigned long addr, unsigned long end,
+static int walk_pud_level(p4d_t *p4d, unsigned long addr, unsigned long end,
 			  unsigned long flags)
 {
 	unsigned long next;
 	pud_t *pudp;
 	int rc = 0;
 
-	pudp = pud_offset(pgd, addr);
+	pudp = pud_offset(p4d, addr);
 	do {
 		if (pud_none(*pudp))
 			return -EINVAL;
@@ -259,6 +259,26 @@ static int walk_pud_level(pgd_t *pgd, unsigned long addr, unsigned long end,
 	return rc;
 }
 
+static int walk_p4d_level(pgd_t *pgd, unsigned long addr, unsigned long end,
+			  unsigned long flags)
+{
+	unsigned long next;
+	p4d_t *p4dp;
+	int rc = 0;
+
+	p4dp = p4d_offset(pgd, addr);
+	do {
+		if (p4d_none(*p4dp))
+			return -EINVAL;
+		next = p4d_addr_end(addr, end);
+		rc = walk_pud_level(p4dp, addr, next, flags);
+		p4dp++;
+		addr = next;
+		cond_resched();
+	} while (addr < end && !rc);
+	return rc;
+}
+
 static DEFINE_MUTEX(cpa_mutex);
 
 static int change_page_attr(unsigned long addr, unsigned long end,
@@ -278,7 +298,7 @@ static int change_page_attr(unsigned long addr, unsigned long end,
 		if (pgd_none(*pgdp))
 			break;
 		next = pgd_addr_end(addr, end);
-		rc = walk_pud_level(pgdp, addr, next, flags);
+		rc = walk_p4d_level(pgdp, addr, next, flags);
 		if (rc)
 			break;
 		cond_resched();
@@ -319,6 +339,7 @@ void __kernel_map_pages(struct page *page, int numpages, int enable)
 	unsigned long address;
 	int nr, i, j;
 	pgd_t *pgd;
+	p4d_t *p4d;
 	pud_t *pud;
 	pmd_t *pmd;
 	pte_t *pte;
@@ -326,7 +347,8 @@ void __kernel_map_pages(struct page *page, int numpages, int enable)
 	for (i = 0; i < numpages;) {
 		address = page_to_phys(page + i);
 		pgd = pgd_offset_k(address);
-		pud = pud_offset(pgd, address);
+		p4d = p4d_offset(pgd, address);
+		pud = pud_offset(p4d, address);
 		pmd = pmd_offset(pud, address);
 		pte = pte_offset_kernel(pmd, address);
 		nr = (unsigned long)pte >> ilog2(sizeof(long));
diff --git a/arch/s390/mm/pgalloc.c b/arch/s390/mm/pgalloc.c
index f502cbe657af..18918e394ce4 100644
--- a/arch/s390/mm/pgalloc.c
+++ b/arch/s390/mm/pgalloc.c
@@ -76,29 +76,46 @@ static void __crst_table_upgrade(void *arg)
 	__tlb_flush_local();
 }
 
-int crst_table_upgrade(struct mm_struct *mm)
+int crst_table_upgrade(struct mm_struct *mm, unsigned long end)
 {
 	unsigned long *table, *pgd;
+	int rc, notify;
 
-	/* upgrade should only happen from 3 to 4 levels */
-	BUG_ON(mm->context.asce_limit != (1UL << 42));
-
-	table = crst_table_alloc(mm);
-	if (!table)
+	/* upgrade should only happen from 3 to 4, 3 to 5, or 4 to 5 levels */
+	BUG_ON(mm->context.asce_limit < (1UL << 42));
+	if (end >= TASK_SIZE_MAX)
 		return -ENOMEM;
-
-	spin_lock_bh(&mm->page_table_lock);
-	pgd = (unsigned long *) mm->pgd;
-	crst_table_init(table, _REGION2_ENTRY_EMPTY);
-	pgd_populate(mm, (pgd_t *) table, (pud_t *) pgd);
-	mm->pgd = (pgd_t *) table;
-	mm->context.asce_limit = 1UL << 53;
-	mm->context.asce = __pa(mm->pgd) | _ASCE_TABLE_LENGTH |
-			   _ASCE_USER_BITS | _ASCE_TYPE_REGION2;
-	spin_unlock_bh(&mm->page_table_lock);
-
-	on_each_cpu(__crst_table_upgrade, mm, 0);
-	return 0;
+	rc = 0;
+	notify = 0;
+	while (mm->context.asce_limit < end) {
+		table = crst_table_alloc(mm);
+		if (!table) {
+			rc = -ENOMEM;
+			break;
+		}
+		spin_lock_bh(&mm->page_table_lock);
+		pgd = (unsigned long *) mm->pgd;
+		if (mm->context.asce_limit == (1UL << 42)) {
+			crst_table_init(table, _REGION2_ENTRY_EMPTY);
+			p4d_populate(mm, (p4d_t *) table, (pud_t *) pgd);
+			mm->pgd = (pgd_t *) table;
+			mm->context.asce_limit = 1UL << 53;
+			mm->context.asce = __pa(mm->pgd) | _ASCE_TABLE_LENGTH |
+				_ASCE_USER_BITS | _ASCE_TYPE_REGION2;
+		} else {
+			crst_table_init(table, _REGION1_ENTRY_EMPTY);
+			pgd_populate(mm, (pgd_t *) table, (p4d_t *) pgd);
+			mm->pgd = (pgd_t *) table;
+			mm->context.asce_limit = -PAGE_SIZE;
+			mm->context.asce = __pa(mm->pgd) | _ASCE_TABLE_LENGTH |
+				_ASCE_USER_BITS | _ASCE_TYPE_REGION1;
+		}
+		notify = 1;
+		spin_unlock_bh(&mm->page_table_lock);
+	}
+	if (notify)
+		on_each_cpu(__crst_table_upgrade, mm, 0);
+	return rc;
 }
 
 void crst_table_downgrade(struct mm_struct *mm)
@@ -274,7 +291,7 @@ static void __tlb_remove_table(void *_table)
 	struct page *page = pfn_to_page(__pa(table) >> PAGE_SHIFT);
 
 	switch (mask) {
-	case 0:		/* pmd or pud */
+	case 0:		/* pmd, pud, or p4d */
 		free_pages((unsigned long) table, 2);
 		break;
 	case 1:		/* lower 2K of a 4K page table */
diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c
index 947b66a5cdba..d4d409ba206b 100644
--- a/arch/s390/mm/pgtable.c
+++ b/arch/s390/mm/pgtable.c
@@ -610,6 +610,7 @@ bool test_and_clear_guest_dirty(struct mm_struct *mm, unsigned long addr)
 {
 	spinlock_t *ptl;
 	pgd_t *pgd;
+	p4d_t *p4d;
 	pud_t *pud;
 	pmd_t *pmd;
 	pgste_t pgste;
@@ -618,7 +619,10 @@ bool test_and_clear_guest_dirty(struct mm_struct *mm, unsigned long addr)
 	bool dirty;
 
 	pgd = pgd_offset(mm, addr);
-	pud = pud_alloc(mm, pgd, addr);
+	p4d = p4d_alloc(mm, pgd, addr);
+	if (!p4d)
+		return false;
+	pud = pud_alloc(mm, p4d, addr);
 	if (!pud)
 		return false;
 	pmd = pmd_alloc(mm, pud, addr);
diff --git a/arch/s390/mm/vmem.c b/arch/s390/mm/vmem.c
index c33c94b4be60..d8398962a723 100644
--- a/arch/s390/mm/vmem.c
+++ b/arch/s390/mm/vmem.c
@@ -38,6 +38,17 @@ static void __ref *vmem_alloc_pages(unsigned int order)
 	return (void *) memblock_alloc(size, size);
 }
 
+static inline p4d_t *vmem_p4d_alloc(void)
+{
+	p4d_t *p4d = NULL;
+
+	p4d = vmem_alloc_pages(2);
+	if (!p4d)
+		return NULL;
+	clear_table((unsigned long *) p4d, _REGION2_ENTRY_EMPTY, PAGE_SIZE * 4);
+	return p4d;
+}
+
 static inline pud_t *vmem_pud_alloc(void)
 {
 	pud_t *pud = NULL;
@@ -85,6 +96,7 @@ static int vmem_add_mem(unsigned long start, unsigned long size)
 	unsigned long end = start + size;
 	unsigned long address = start;
 	pgd_t *pg_dir;
+	p4d_t *p4_dir;
 	pud_t *pu_dir;
 	pmd_t *pm_dir;
 	pte_t *pt_dir;
@@ -102,12 +114,19 @@ static int vmem_add_mem(unsigned long start, unsigned long size)
 	while (address < end) {
 		pg_dir = pgd_offset_k(address);
 		if (pgd_none(*pg_dir)) {
+			p4_dir = vmem_p4d_alloc();
+			if (!p4_dir)
+				goto out;
+			pgd_populate(&init_mm, pg_dir, p4_dir);
+		}
+		p4_dir = p4d_offset(pg_dir, address);
+		if (p4d_none(*p4_dir)) {
 			pu_dir = vmem_pud_alloc();
 			if (!pu_dir)
 				goto out;
-			pgd_populate(&init_mm, pg_dir, pu_dir);
+			p4d_populate(&init_mm, p4_dir, pu_dir);
 		}
-		pu_dir = pud_offset(pg_dir, address);
+		pu_dir = pud_offset(p4_dir, address);
 		if (MACHINE_HAS_EDAT2 && pud_none(*pu_dir) && address &&
 		    !(address & ~PUD_MASK) && (address + PUD_SIZE <= end) &&
 		     !debug_pagealloc_enabled()) {
@@ -161,6 +180,7 @@ static void vmem_remove_range(unsigned long start, unsigned long size)
 	unsigned long end = start + size;
 	unsigned long address = start;
 	pgd_t *pg_dir;
+	p4d_t *p4_dir;
 	pud_t *pu_dir;
 	pmd_t *pm_dir;
 	pte_t *pt_dir;
@@ -172,7 +192,12 @@ static void vmem_remove_range(unsigned long start, unsigned long size)
 			address += PGDIR_SIZE;
 			continue;
 		}
-		pu_dir = pud_offset(pg_dir, address);
+		p4_dir = p4d_offset(pg_dir, address);
+		if (p4d_none(*p4_dir)) {
+			address += P4D_SIZE;
+			continue;
+		}
+		pu_dir = pud_offset(p4_dir, address);
 		if (pud_none(*pu_dir)) {
 			address += PUD_SIZE;
 			continue;
@@ -213,6 +238,7 @@ int __meminit vmemmap_populate(unsigned long start, unsigned long end, int node)
 	unsigned long pgt_prot, sgt_prot;
 	unsigned long address = start;
 	pgd_t *pg_dir;
+	p4d_t *p4_dir;
 	pud_t *pu_dir;
 	pmd_t *pm_dir;
 	pte_t *pt_dir;
@@ -227,13 +253,21 @@ int __meminit vmemmap_populate(unsigned long start, unsigned long end, int node)
 	for (address = start; address < end;) {
 		pg_dir = pgd_offset_k(address);
 		if (pgd_none(*pg_dir)) {
+			p4_dir = vmem_p4d_alloc();
+			if (!p4_dir)
+				goto out;
+			pgd_populate(&init_mm, pg_dir, p4_dir);
+		}
+
+		p4_dir = p4d_offset(pg_dir, address);
+		if (p4d_none(*p4_dir)) {
 			pu_dir = vmem_pud_alloc();
 			if (!pu_dir)
 				goto out;
-			pgd_populate(&init_mm, pg_dir, pu_dir);
+			p4d_populate(&init_mm, p4_dir, pu_dir);
 		}
 
-		pu_dir = pud_offset(pg_dir, address);
+		pu_dir = pud_offset(p4_dir, address);
 		if (pud_none(*pu_dir)) {
 			pm_dir = vmem_pmd_alloc();
 			if (!pm_dir)
-- 
cgit 


From 92acfb7406171038ae4e3b6041576642cb75b529 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Tue, 9 May 2017 11:01:26 +0200
Subject: s390: add missing header includes for type checking

Add missing include statements to make sure that prototypes match
implementation. As reported by sparse:

arch/s390/crypto/arch_random.c:18:1:
  warning: symbol 's390_arch_random_available' was not declared. Should it be static?
arch/s390/kernel/traps.c:279:13: warning:
  symbol 'trap_init' was not declared. Should it be static?

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/crypto/arch_random.c | 1 +
 arch/s390/kernel/traps.c       | 1 +
 2 files changed, 2 insertions(+)

(limited to 'arch/s390')

diff --git a/arch/s390/crypto/arch_random.c b/arch/s390/crypto/arch_random.c
index 9317b3e645e2..36aefc07d10c 100644
--- a/arch/s390/crypto/arch_random.c
+++ b/arch/s390/crypto/arch_random.c
@@ -12,6 +12,7 @@
 
 #include <linux/kernel.h>
 #include <linux/atomic.h>
+#include <linux/random.h>
 #include <linux/static_key.h>
 #include <asm/cpacf.h>
 
diff --git a/arch/s390/kernel/traps.c b/arch/s390/kernel/traps.c
index f787b9d8f54c..442e5423ce3d 100644
--- a/arch/s390/kernel/traps.c
+++ b/arch/s390/kernel/traps.c
@@ -21,6 +21,7 @@
 #include <linux/mm.h>
 #include <linux/slab.h>
 #include <linux/uaccess.h>
+#include <linux/cpu.h>
 #include <asm/fpu/api.h>
 #include "entry.h"
 
-- 
cgit 


From d12a3d603690ba84b7e3fd357f05b96e777d4630 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Tue, 9 May 2017 13:44:43 +0200
Subject: s390/mm: add __rcu annotations

Add __rcu annotations so sparse correctly warns only if "slot" gets
derefenced without using rcu_dereference(). Right now we get warnings
because of the missing annotation:

arch/s390/mm/gmap.c:135:17: warning: incorrect type in assignment (different address spaces)
arch/s390/mm/gmap.c:135:17:    expected void **slot
arch/s390/mm/gmap.c:135:17:    got void [noderef] <asn:4>**

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/mm/gmap.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/mm/gmap.c b/arch/s390/mm/gmap.c
index fbd664e48098..4fb3d3cdb370 100644
--- a/arch/s390/mm/gmap.c
+++ b/arch/s390/mm/gmap.c
@@ -125,7 +125,7 @@ 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;
+	void __rcu **slot;
 	int i, nr;
 
 	/* A radix tree is freed by deleting all of its entries */
@@ -150,7 +150,7 @@ static void gmap_rmap_radix_tree_free(struct radix_tree_root *root)
 	struct radix_tree_iter iter;
 	unsigned long indices[16];
 	unsigned long index;
-	void **slot;
+	void __rcu **slot;
 	int i, nr;
 
 	/* A radix tree is freed by deleting all of its entries */
@@ -1011,7 +1011,7 @@ EXPORT_SYMBOL_GPL(gmap_read_table);
 static inline void gmap_insert_rmap(struct gmap *sg, unsigned long vmaddr,
 				    struct gmap_rmap *rmap)
 {
-	void **slot;
+	void __rcu **slot;
 
 	BUG_ON(!gmap_is_shadow(sg));
 	slot = radix_tree_lookup_slot(&sg->host_to_rmap, vmaddr >> PAGE_SHIFT);
-- 
cgit 


From 6c386da799078fddb482bc5c1ab6a62c9a360e4e Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Fri, 12 May 2017 12:57:40 +0200
Subject: s390: use two more generic header files

I missed at least these two header files where we can make use of the
generic ones.  vga.h is another one, however that is already addressed
by a patch from Jiri Slaby.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/include/asm/Kbuild   |  2 ++
 arch/s390/include/asm/device.h | 10 ----------
 arch/s390/include/asm/fb.h     | 12 ------------
 3 files changed, 2 insertions(+), 22 deletions(-)
 delete mode 100644 arch/s390/include/asm/device.h
 delete mode 100644 arch/s390/include/asm/fb.h

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/Kbuild b/arch/s390/include/asm/Kbuild
index 45092b12f54f..b3c88479feba 100644
--- a/arch/s390/include/asm/Kbuild
+++ b/arch/s390/include/asm/Kbuild
@@ -1,10 +1,12 @@
 generic-y += asm-offsets.h
 generic-y += cacheflush.h
 generic-y += clkdev.h
+generic-y += device.h
 generic-y += dma-contiguous.h
 generic-y += div64.h
 generic-y += emergency-restart.h
 generic-y += export.h
+generic-y += fb.h
 generic-y += irq_regs.h
 generic-y += irq_work.h
 generic-y += kmap_types.h
diff --git a/arch/s390/include/asm/device.h b/arch/s390/include/asm/device.h
deleted file mode 100644
index 5203fc87f080..000000000000
--- a/arch/s390/include/asm/device.h
+++ /dev/null
@@ -1,10 +0,0 @@
-/*
- * Arch specific extensions to struct device
- *
- * This file is released under the GPLv2
- */
-struct dev_archdata {
-};
-
-struct pdev_archdata {
-};
diff --git a/arch/s390/include/asm/fb.h b/arch/s390/include/asm/fb.h
deleted file mode 100644
index c7df38030992..000000000000
--- a/arch/s390/include/asm/fb.h
+++ /dev/null
@@ -1,12 +0,0 @@
-#ifndef _ASM_FB_H_
-#define _ASM_FB_H_
-#include <linux/fb.h>
-
-#define fb_pgprotect(...) do {} while (0)
-
-static inline int fb_is_primary_device(struct fb_info *info)
-{
-	return 0;
-}
-
-#endif /* _ASM_FB_H_ */
-- 
cgit 


From b29e061bb7a8d2e2fbf7826fb30ab5477d595e92 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Fri, 12 May 2017 15:31:38 +0200
Subject: s390/cputime: simplify account_system_index_scaled

The account_system_index_scaled gets two cputime values, a raw value
derived from CPU timer deltas and a scaled value. The scaled value
is always calculated from the raw value, the code can be simplified
by moving the scale_vtime call into account_system_index_scaled.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/vtime.c | 14 +++++---------
 1 file changed, 5 insertions(+), 9 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c
index 072d84ba42a3..dd7178fbb4f3 100644
--- a/arch/s390/kernel/vtime.c
+++ b/arch/s390/kernel/vtime.c
@@ -110,11 +110,10 @@ static inline u64 scale_vtime(u64 vtime)
 	return vtime;
 }
 
-static void account_system_index_scaled(struct task_struct *p,
-					u64 cputime, u64 scaled,
+static void account_system_index_scaled(struct task_struct *p, u64 cputime,
 					enum cpu_usage_stat index)
 {
-	p->stimescaled += cputime_to_nsecs(scaled);
+	p->stimescaled += cputime_to_nsecs(scale_vtime(cputime));
 	account_system_index_time(p, cputime_to_nsecs(cputime), index);
 }
 
@@ -176,14 +175,11 @@ static int do_account_vtime(struct task_struct *tsk)
 	}
 
 	if (system)
-		account_system_index_scaled(tsk, system, scale_vtime(system),
-					    CPUTIME_SYSTEM);
+		account_system_index_scaled(tsk, system, CPUTIME_SYSTEM);
 	if (hardirq)
-		account_system_index_scaled(tsk, hardirq, scale_vtime(hardirq),
-					    CPUTIME_IRQ);
+		account_system_index_scaled(tsk, hardirq, CPUTIME_IRQ);
 	if (softirq)
-		account_system_index_scaled(tsk, softirq, scale_vtime(softirq),
-					    CPUTIME_SOFTIRQ);
+		account_system_index_scaled(tsk, softirq, CPUTIME_SOFTIRQ);
 
 	steal = S390_lowcore.steal_timer;
 	if ((s64) steal > 0) {
-- 
cgit 


From 35bb092a91b4c43a2079d3d5458b74a059455b6c Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Mon, 15 May 2017 10:23:38 +0200
Subject: s390/vdso: use _install_special_mapping to establish vdso

Switch to the improved _install_special_mapping function to install
the vdso mapping. This has two advantages, the arch_vma_name function
is not needed anymore and the vdso vma still has its name after its
memory location has been changed with mremap.

Tested-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/vdso.c | 91 ++++++++++++++++++++++++++++++++++---------------
 1 file changed, 64 insertions(+), 27 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/kernel/vdso.c b/arch/s390/kernel/vdso.c
index 10516ae3b55e..b89d19f6f2ab 100644
--- a/arch/s390/kernel/vdso.c
+++ b/arch/s390/kernel/vdso.c
@@ -50,6 +50,56 @@ static struct page **vdso64_pagelist;
  */
 unsigned int __read_mostly vdso_enabled = 1;
 
+static int vdso_fault(const struct vm_special_mapping *sm,
+		      struct vm_area_struct *vma, struct vm_fault *vmf)
+{
+	struct page **vdso_pagelist;
+	unsigned long vdso_pages;
+
+	vdso_pagelist = vdso64_pagelist;
+	vdso_pages = vdso64_pages;
+#ifdef CONFIG_COMPAT
+	if (is_compat_task()) {
+		vdso_pagelist = vdso32_pagelist;
+		vdso_pages = vdso32_pages;
+	}
+#endif
+
+	if (vmf->pgoff >= vdso_pages)
+		return VM_FAULT_SIGBUS;
+
+	vmf->page = vdso_pagelist[vmf->pgoff];
+	get_page(vmf->page);
+	return 0;
+}
+
+static int vdso_mremap(const struct vm_special_mapping *sm,
+		       struct vm_area_struct *vma)
+{
+	unsigned long vdso_pages;
+
+	vdso_pages = vdso64_pages;
+#ifdef CONFIG_COMPAT
+	if (is_compat_task())
+		vdso_pages = vdso32_pages;
+#endif
+
+	if ((vdso_pages << PAGE_SHIFT) != vma->vm_end - vma->vm_start)
+		return -EINVAL;
+
+	if (WARN_ON_ONCE(current->mm != vma->vm_mm))
+		return -EFAULT;
+
+	current->mm->context.vdso_base = vma->vm_start;
+	return 0;
+}
+
+static const struct vm_special_mapping vdso_mapping = {
+	.name = "[vdso]",
+	.fault = vdso_fault,
+	.mremap = vdso_mremap,
+};
+
 static int __init vdso_setup(char *s)
 {
 	unsigned long val;
@@ -181,7 +231,7 @@ static void vdso_init_cr5(void)
 int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
 {
 	struct mm_struct *mm = current->mm;
-	struct page **vdso_pagelist;
+	struct vm_area_struct *vma;
 	unsigned long vdso_pages;
 	unsigned long vdso_base;
 	int rc;
@@ -194,13 +244,10 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
 	if (!uses_interp)
 		return 0;
 
-	vdso_pagelist = vdso64_pagelist;
 	vdso_pages = vdso64_pages;
 #ifdef CONFIG_COMPAT
-	if (is_compat_task()) {
-		vdso_pagelist = vdso32_pagelist;
+	if (is_compat_task())
 		vdso_pages = vdso32_pages;
-	}
 #endif
 	/*
 	 * vDSO has a problem and was disabled, just don't "enable" it for
@@ -209,8 +256,6 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
 	if (vdso_pages == 0)
 		return 0;
 
-	current->mm->context.vdso_base = 0;
-
 	/*
 	 * pick a base address for the vDSO in process space. We try to put
 	 * it at vdso_base which is the "natural" base for it, but we might
@@ -224,13 +269,6 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
 		goto out_up;
 	}
 
-	/*
-	 * Put vDSO base into mm struct. We need to do this before calling
-	 * install_special_mapping or the perf counter mmap tracking code
-	 * will fail to recognise it as a vDSO (since arch_vma_name fails).
-	 */
-	current->mm->context.vdso_base = vdso_base;
-
 	/*
 	 * our vma flags don't have VM_WRITE so by default, the process
 	 * isn't allowed to write those pages.
@@ -241,24 +279,23 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
 	 * It's fine to use that for setting breakpoints in the vDSO code
 	 * pages though.
 	 */
-	rc = install_special_mapping(mm, vdso_base, vdso_pages << PAGE_SHIFT,
-				     VM_READ|VM_EXEC|
-				     VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC,
-				     vdso_pagelist);
-	if (rc)
-		current->mm->context.vdso_base = 0;
+	vma = _install_special_mapping(mm, vdso_base, vdso_pages << PAGE_SHIFT,
+				       VM_READ|VM_EXEC|
+				       VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC,
+				       &vdso_mapping);
+	if (IS_ERR(vma)) {
+		rc = PTR_ERR(vma);
+		goto out_up;
+	}
+
+	current->mm->context.vdso_base = vdso_base;
+	rc = 0;
+
 out_up:
 	up_write(&mm->mmap_sem);
 	return rc;
 }
 
-const char *arch_vma_name(struct vm_area_struct *vma)
-{
-	if (vma->vm_mm && vma->vm_start == vma->vm_mm->context.vdso_base)
-		return "[vdso]";
-	return NULL;
-}
-
 static int __init vdso_init(void)
 {
 	int i;
-- 
cgit 


From c4684f98d3453dd07cc7ce67e0e795330eeec9c5 Mon Sep 17 00:00:00 2001
From: Harald Freudenberger <freude@linux.vnet.ibm.com>
Date: Thu, 11 May 2017 17:15:54 +0200
Subject: s390/crypto: fix aes/paes Kconfig dependeny

The s390_paes and the s390_aes kernel module used just one
config symbol CONFIG_CRYPTO_AES. As paes has a dependency
to PKEY and this requires ZCRYPT the aes module also had
a dependency to the zcrypt device driver which is not true.
Fixed by introducing a new config symbol CONFIG_CRYPTO_PAES
which has dependencies to PKEY and ZCRYPT. Removed the
dependency for the aes module to ZCRYPT.

Signed-off-by: Harald Freudenberger <freude@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/crypto/Makefile | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

(limited to 'arch/s390')

diff --git a/arch/s390/crypto/Makefile b/arch/s390/crypto/Makefile
index 678d9863e3f0..ad4bd777768d 100644
--- a/arch/s390/crypto/Makefile
+++ b/arch/s390/crypto/Makefile
@@ -6,7 +6,8 @@ obj-$(CONFIG_CRYPTO_SHA1_S390) += sha1_s390.o sha_common.o
 obj-$(CONFIG_CRYPTO_SHA256_S390) += sha256_s390.o sha_common.o
 obj-$(CONFIG_CRYPTO_SHA512_S390) += sha512_s390.o sha_common.o
 obj-$(CONFIG_CRYPTO_DES_S390) += des_s390.o
-obj-$(CONFIG_CRYPTO_AES_S390) += aes_s390.o paes_s390.o
+obj-$(CONFIG_CRYPTO_AES_S390) += aes_s390.o
+obj-$(CONFIG_CRYPTO_PAES_S390) += paes_s390.o
 obj-$(CONFIG_S390_PRNG) += prng.o
 obj-$(CONFIG_CRYPTO_GHASH_S390) += ghash_s390.o
 obj-$(CONFIG_CRYPTO_CRC32_S390) += crc32-vx_s390.o
-- 
cgit 


From 9cf8edb7a33b7689d4c3429b1d339d29e7c3db86 Mon Sep 17 00:00:00 2001
From: Christian Borntraeger <borntraeger@de.ibm.com>
Date: Thu, 18 May 2017 13:04:16 +0200
Subject: s390/smp: fix false positive kmemleak of mcesa data structure

I get number of CPUs - 1 kmemleak hits like

unreferenced object 0x37ec6f000 (size 1024):
  comm "swapper/0", pid 1, jiffies 4294937330 (age 889.690s)
  hex dump (first 32 bytes):
    6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b  kkkkkkkkkkkkkkkk
    6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b  kkkkkkkkkkkkkkkk
  backtrace:
    [<000000000034a848>] kmem_cache_alloc+0x2b8/0x3d0
    [<00000000001164de>] __cpu_up+0x456/0x488
    [<000000000016f60c>] bringup_cpu+0x4c/0xd0
    [<000000000016d5d2>] cpuhp_invoke_callback+0xe2/0x9e8
    [<000000000016f3c6>] cpuhp_up_callbacks+0x5e/0x110
    [<000000000016f988>] _cpu_up+0xe0/0x158
    [<000000000016faf0>] do_cpu_up+0xf0/0x110
    [<0000000000dae1ee>] smp_init+0x126/0x130
    [<0000000000d9bd04>] kernel_init_freeable+0x174/0x2e0
    [<000000000089fc62>] kernel_init+0x2a/0x148
    [<00000000008adce2>] kernel_thread_starter+0x6/0xc
    [<00000000008adcdc>] kernel_thread_starter+0x0/0xc
    [<ffffffffffffffff>] 0xffffffffffffffff

The pointer of this data structure is stored in the prefix page of that
CPU together with some extra bits ORed into the the low bits.
Mark the data structure as non-leak.

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/smp.c | 3 +++
 1 file changed, 3 insertions(+)

(limited to 'arch/s390')

diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c
index 363000a77ffc..1020a11a24e5 100644
--- a/arch/s390/kernel/smp.c
+++ b/arch/s390/kernel/smp.c
@@ -26,6 +26,7 @@
 #include <linux/err.h>
 #include <linux/spinlock.h>
 #include <linux/kernel_stat.h>
+#include <linux/kmemleak.h>
 #include <linux/delay.h>
 #include <linux/interrupt.h>
 #include <linux/irqflags.h>
@@ -207,6 +208,8 @@ static int pcpu_alloc_lowcore(struct pcpu *pcpu, int cpu)
 				kmem_cache_alloc(pcpu_mcesa_cache, GFP_KERNEL);
 			if (!mcesa_origin)
 				goto out;
+			/* The pointer is stored with mcesa_bits ORed in */
+			kmemleak_not_leak((void *) mcesa_origin);
 			mcesa_bits = MACHINE_HAS_GS ? 11 : 0;
 		}
 	} else {
-- 
cgit 


From bf10b6687c19484cdf28107ec9b4be51c06c7746 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Fri, 19 May 2017 09:07:23 +0200
Subject: s390/smp: use sigp condition code define

Use proper define instead of open-coding the condition code value.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/include/asm/sigp.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/sigp.h b/arch/s390/include/asm/sigp.h
index 72df5f2de6b0..020a8814d511 100644
--- a/arch/s390/include/asm/sigp.h
+++ b/arch/s390/include/asm/sigp.h
@@ -59,7 +59,7 @@ static inline int __pcpu_sigp(u16 addr, u8 order, unsigned long parm,
 	int cc;
 
 	cc = ____pcpu_sigp(addr, order, parm, &_status);
-	if (status && cc == 1)
+	if (status && cc == SIGP_CC_STATUS_STORED)
 		*status = _status;
 	return cc;
 }
-- 
cgit 


From fe7b274729fc0bab9b4238f875695d36726a6b10 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Mon, 22 May 2017 13:16:00 +0200
Subject: s390/fault: use _ASCE_ORIGIN instead of PAGE_MASK

When masking an ASCE to get its origin use the corresponding define
instead of the unrelated PAGE_MASK.
This doesn't fix a bug since both masks are identical.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/mm/fault.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'arch/s390')

diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c
index 5845d3028ffc..14f25798b001 100644
--- a/arch/s390/mm/fault.c
+++ b/arch/s390/mm/fault.c
@@ -130,7 +130,7 @@ static int bad_address(void *p)
 
 static void dump_pagetable(unsigned long asce, unsigned long address)
 {
-	unsigned long *table = __va(asce & PAGE_MASK);
+	unsigned long *table = __va(asce & _ASCE_ORIGIN);
 
 	pr_alert("AS:%016lx ", asce);
 	switch (asce & _ASCE_TYPE_MASK) {
-- 
cgit 


From 8aa8680aa383bf6e2ac2a7d3369097268c75f7da Mon Sep 17 00:00:00 2001
From: Thomas Huth <thuth@redhat.com>
Date: Mon, 22 May 2017 17:20:53 +0200
Subject: s390: Remove 'message security assist' from the list of vital
 facilities

The code in arch/s390/crypto checks for the availability of the
'message security assist' facility on its own, either by using
module_cpu_feature_match(MSA, ...) or by checking the facility
bit during cpacf_query(). Thus setting the MSA facility bit in
gen_facilities.c as hard requirement is not necessary. We can
remove it here, so that the kernel can also run on systems that
do not provide the MSA facility yet (like the emulated environment
of QEMU, for example).

Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/tools/gen_facilities.c | 1 -
 1 file changed, 1 deletion(-)

(limited to 'arch/s390')

diff --git a/arch/s390/tools/gen_facilities.c b/arch/s390/tools/gen_facilities.c
index be63fbd699fd..cfe3f76bc2de 100644
--- a/arch/s390/tools/gen_facilities.c
+++ b/arch/s390/tools/gen_facilities.c
@@ -35,7 +35,6 @@ static struct facility_def facility_defs[] = {
 #endif
 #ifdef CONFIG_HAVE_MARCH_Z9_109_FEATURES
 			7,  /* stfle */
-			17, /* message security assist */
 			21, /* extended-immediate facility */
 			25, /* store clock fast */
 #endif
-- 
cgit 


From 53d7f25f09eb0ea7cb119d86590f3992656a6892 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Tue, 23 May 2017 09:17:24 +0200
Subject: s390/facilities: remove stfle requirement

All call sites of "stfle" check if the instruction is available before
executing it. Therefore there is no reason to have the corresponding
facility bit set within the architecture level set.
This removes the last more or less odd bit from the list.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/tools/gen_facilities.c | 1 -
 1 file changed, 1 deletion(-)

(limited to 'arch/s390')

diff --git a/arch/s390/tools/gen_facilities.c b/arch/s390/tools/gen_facilities.c
index cfe3f76bc2de..025ea20fc4b4 100644
--- a/arch/s390/tools/gen_facilities.c
+++ b/arch/s390/tools/gen_facilities.c
@@ -34,7 +34,6 @@ static struct facility_def facility_defs[] = {
 			18, /* long displacement facility */
 #endif
 #ifdef CONFIG_HAVE_MARCH_Z9_109_FEATURES
-			7,  /* stfle */
 			21, /* extended-immediate facility */
 			25, /* store clock fast */
 #endif
-- 
cgit 


From f5bbd7219891364a6ba1231cb751905045fd4466 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Fri, 19 May 2017 15:32:09 +0200
Subject: s390/ptrace: guarded storage regset for the current task

The regset functions for guarded storage are supposed to work on
the current task as well. For task == current add the required
load and store instructions for the guarded storage control block.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/ptrace.c | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c
index 488c5bb8dc77..252ed61a128b 100644
--- a/arch/s390/kernel/ptrace.c
+++ b/arch/s390/kernel/ptrace.c
@@ -1160,6 +1160,8 @@ static int s390_gs_cb_get(struct task_struct *target,
 		return -ENODEV;
 	if (!data)
 		return -ENODATA;
+	if (target == current)
+		save_gs_cb(data);
 	return user_regset_copyout(&pos, &count, &kbuf, &ubuf,
 				   data, 0, sizeof(struct gs_cb));
 }
@@ -1170,6 +1172,7 @@ static int s390_gs_cb_set(struct task_struct *target,
 			  const void *kbuf, const void __user *ubuf)
 {
 	struct gs_cb *data = target->thread.gs_cb;
+	int rc;
 
 	if (!MACHINE_HAS_GS)
 		return -ENODEV;
@@ -1177,10 +1180,18 @@ static int s390_gs_cb_set(struct task_struct *target,
 		data = kzalloc(sizeof(*data), GFP_KERNEL);
 		if (!data)
 			return -ENOMEM;
+		data->gsd = 25;
 		target->thread.gs_cb = data;
+		if (target == current)
+			__ctl_set_bit(2, 4);
+	} else if (target == current) {
+		save_gs_cb(data);
 	}
-	return user_regset_copyin(&pos, &count, &kbuf, &ubuf,
-				  data, 0, sizeof(struct gs_cb));
+	rc = user_regset_copyin(&pos, &count, &kbuf, &ubuf,
+				data, 0, sizeof(struct gs_cb));
+	if (target == current)
+		restore_gs_cb(data);
+	return rc;
 }
 
 static int s390_gs_bc_get(struct task_struct *target,
-- 
cgit 


From f96c6f72bc792e7add27e40ec859a8f5ea72e72b Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Mon, 22 May 2017 13:27:34 +0200
Subject: s390/mm: remove incorrect _REGION3_ENTRY_ORIGIN define

_REGION3_ENTRY_ORIGIN defines a wrong mask which can be used to
extract a segment table origin from a region 3 table entry. It removes
only the lower 11 instead of 12 bits from a region 3 table entry.
Luckily this bit is currently always zero, so nothing bad happened yet.

In order to avoid future bugs just remove the region 3 specific mask
and use the correct generic _REGION_ENTRY_ORIGIN mask.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/include/asm/pgtable.h | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h
index 3effb26f0e1a..6f8a67a103be 100644
--- a/arch/s390/include/asm/pgtable.h
+++ b/arch/s390/include/asm/pgtable.h
@@ -301,8 +301,6 @@ static inline int is_module_addr(void *addr)
 #define _REGION3_ENTRY_EMPTY	(_REGION_ENTRY_TYPE_R3 | _REGION_ENTRY_INVALID)
 
 #define _REGION3_ENTRY_ORIGIN_LARGE ~0x7fffffffUL /* large page address	     */
-#define _REGION3_ENTRY_ORIGIN  ~0x7ffUL/* region third table origin	     */
-
 #define _REGION3_ENTRY_DIRTY	0x2000	/* SW region dirty bit */
 #define _REGION3_ENTRY_YOUNG	0x1000	/* SW region young bit */
 #define _REGION3_ENTRY_LARGE	0x0400	/* RTTE-format control, large page  */
@@ -641,7 +639,7 @@ static inline unsigned long pud_pfn(pud_t pud)
 {
 	unsigned long origin_mask;
 
-	origin_mask = _REGION3_ENTRY_ORIGIN;
+	origin_mask = _REGION_ENTRY_ORIGIN;
 	if (pud_large(pud))
 		origin_mask = _REGION3_ENTRY_ORIGIN_LARGE;
 	return (pud_val(pud) & origin_mask) >> PAGE_SHIFT;
-- 
cgit 


From cc18b460dc72fc9020edcd617b3c4b23a577ee19 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Sat, 20 May 2017 11:43:26 +0200
Subject: s390/mm: add p?d_folded() helper functions

Introduce and use p?d_folded() functions to clarify the page table
code a bit more.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/include/asm/pgtable.h | 29 ++++++++++++++++++++++-------
 1 file changed, 22 insertions(+), 7 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h
index 6f8a67a103be..57057fb1cc07 100644
--- a/arch/s390/include/asm/pgtable.h
+++ b/arch/s390/include/asm/pgtable.h
@@ -563,18 +563,23 @@ static inline void crdte(unsigned long old, unsigned long new,
 }
 
 /*
- * pgd/pmd/pte query functions
+ * pgd/p4d/pud/pmd/pte query functions
  */
+static inline int pgd_folded(pgd_t pgd)
+{
+	return (pgd_val(pgd) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R1;
+}
+
 static inline int pgd_present(pgd_t pgd)
 {
-	if ((pgd_val(pgd) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R1)
+	if (pgd_folded(pgd))
 		return 1;
 	return (pgd_val(pgd) & _REGION_ENTRY_ORIGIN) != 0UL;
 }
 
 static inline int pgd_none(pgd_t pgd)
 {
-	if ((pgd_val(pgd) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R1)
+	if (pgd_folded(pgd))
 		return 0;
 	return (pgd_val(pgd) & _REGION_ENTRY_INVALID) != 0UL;
 }
@@ -592,16 +597,21 @@ static inline int pgd_bad(pgd_t pgd)
 	return (pgd_val(pgd) & mask) != 0;
 }
 
+static inline int p4d_folded(p4d_t p4d)
+{
+	return (p4d_val(p4d) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R2;
+}
+
 static inline int p4d_present(p4d_t p4d)
 {
-	if ((p4d_val(p4d) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R2)
+	if (p4d_folded(p4d))
 		return 1;
 	return (p4d_val(p4d) & _REGION_ENTRY_ORIGIN) != 0UL;
 }
 
 static inline int p4d_none(p4d_t p4d)
 {
-	if ((p4d_val(p4d) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R2)
+	if (p4d_folded(p4d))
 		return 0;
 	return p4d_val(p4d) == _REGION2_ENTRY_EMPTY;
 }
@@ -614,16 +624,21 @@ static inline unsigned long p4d_pfn(p4d_t p4d)
 	return (p4d_val(p4d) & origin_mask) >> PAGE_SHIFT;
 }
 
+static inline int pud_folded(pud_t pud)
+{
+	return (pud_val(pud) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R3;
+}
+
 static inline int pud_present(pud_t pud)
 {
-	if ((pud_val(pud) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R3)
+	if (pud_folded(pud))
 		return 1;
 	return (pud_val(pud) & _REGION_ENTRY_ORIGIN) != 0UL;
 }
 
 static inline int pud_none(pud_t pud)
 {
-	if ((pud_val(pud) & _REGION_ENTRY_TYPE_MASK) < _REGION_ENTRY_TYPE_R3)
+	if (pud_folded(pud))
 		return 0;
 	return pud_val(pud) == _REGION3_ENTRY_EMPTY;
 }
-- 
cgit 


From c39457ff1fd58bf5cc466025b1467de29ba0437a Mon Sep 17 00:00:00 2001
From: Thomas Richter <tmricht@linux.vnet.ibm.com>
Date: Tue, 23 May 2017 10:49:35 +0200
Subject: s390/perf: fix null string in perf list pmu command

Command 'perf list pmu' displays events which contain
an invalid string "(null)=xxx", where xxx is the pmu event
name, for example:
   cpum_cf/AES_BLOCKED_CYCLES,(null)=AES_BLOCKED_CYCLES/
This is not correct, the invalid string should not be
displayed at all.

It is caused by an obsolete term in the
sysfs attribute file for each s390 CPUMF counter event.
Reading from the sysfs file also displays the event
name.

Fix this by omitting the event name.  This patch makes
s390 CPUMF sysfs files consistent with other plattforms.

This is an interface change between user and kernel
but does not break anything. Reading from a counter event
sysfs file should only list terms mentioned in the
/sys/bus/event_source/devices/<cpumf>/format directory.
Name is not listed.

Reported-by: Zvonko Kosic <zvonko.kosic@de.ibm.com>
Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Reviewed-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/perf_event.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/kernel/perf_event.c b/arch/s390/kernel/perf_event.c
index 955a7b6fa0a4..93a386f4a3b5 100644
--- a/arch/s390/kernel/perf_event.c
+++ b/arch/s390/kernel/perf_event.c
@@ -245,6 +245,5 @@ ssize_t cpumf_events_sysfs_show(struct device *dev,
 	struct perf_pmu_events_attr *pmu_attr;
 
 	pmu_attr = container_of(attr, struct perf_pmu_events_attr, attr);
-	return sprintf(page, "event=0x%04llx,name=%s\n",
-		       pmu_attr->id, attr->attr.name);
+	return sprintf(page, "event=0x%04llx\n", pmu_attr->id);
 }
-- 
cgit 


From 7e9710af2301e76421a6a8f2655937651279f5c4 Mon Sep 17 00:00:00 2001
From: Logan Gunthorpe <logang@deltatee.com>
Date: Mon, 29 May 2017 13:13:08 -0600
Subject: s390: provide default ioremap and iounmap declaration

Move the CONFIG_PCI device so that ioremap and iounmap are always
available. This looks safe as there's nothing PCI specific in the
implementation of these functions.

I have designs to use these functions in scatterlist.c where they'd likely
never be called without CONFIG_PCI set, but this is needed to compile
such changes.

Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Sebastian Ott <sebott@linux.vnet.ibm.com>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/include/asm/io.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/io.h b/arch/s390/include/asm/io.h
index 437e9af96688..904e4b3af95d 100644
--- a/arch/s390/include/asm/io.h
+++ b/arch/s390/include/asm/io.h
@@ -25,8 +25,6 @@ void unxlate_dev_mem_ptr(phys_addr_t phys, void *addr);
 
 #define IO_SPACE_LIMIT 0
 
-#ifdef CONFIG_PCI
-
 #define ioremap_nocache(addr, size)	ioremap(addr, size)
 #define ioremap_wc			ioremap_nocache
 #define ioremap_wt			ioremap_nocache
@@ -49,6 +47,8 @@ static inline void ioport_unmap(void __iomem *p)
 {
 }
 
+#ifdef CONFIG_PCI
+
 /*
  * s390 needs a private implementation of pci_iomap since ioremap with its
  * offset parameter isn't sufficient. That's because BAR spaces are not
-- 
cgit 


From 2b7b9817c2dbfce0501a313db817718fc5ef6524 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Tue, 6 Jun 2017 10:04:26 +0200
Subject: s390/dumpstack: remove raw stack dump

Remove raw stack dumps that are printed before call traces in case of
a warning, or the 'l' sysrq trigger (show a stack backtrace for all
active CPUs).

Besides that a raw stack dump should not be shown for the 'l' sysrq
trigger the value of the dump is close to zero. That's also why we
don't print it in case of a panic since ages anymore. That this is
still printed on warnings is just a leftover. So get rid of this
completely.

The following won't be printed anymore with this change:

Stack:
       00000000bbc4fbc8 00000000bbc4fc58 0000000000000003 0000000000000000
       00000000bbc4fcf8 00000000bbc4fc70 00000000bbc4fc70 0000000000000020
       000000007fe00098 00000000bfe8be00 00000000bbc4fe94 000000000000000a
       000000000000000c 00000000bbc4fcc0 0000000000000000 0000000000000000
       000000000095b930 0000000000113366 00000000bbc4fc58 00000000bbc4fca0

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/dumpstack.c | 29 ++++-------------------------
 1 file changed, 4 insertions(+), 25 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/kernel/dumpstack.c b/arch/s390/kernel/dumpstack.c
index 829e1c53005c..62904890d127 100644
--- a/arch/s390/kernel/dumpstack.c
+++ b/arch/s390/kernel/dumpstack.c
@@ -98,8 +98,10 @@ static int show_address(void *data, unsigned long address, int reliable)
 	return 0;
 }
 
-static void show_trace(struct task_struct *task, unsigned long sp)
+void show_stack(struct task_struct *task, unsigned long *stack)
 {
+	unsigned long sp = (unsigned long) stack;
+
 	if (!sp)
 		sp = task ? task->thread.ksp : current_stack_pointer();
 	printk("Call Trace:\n");
@@ -109,29 +111,6 @@ static void show_trace(struct task_struct *task, unsigned long sp)
 	debug_show_held_locks(task);
 }
 
-void show_stack(struct task_struct *task, unsigned long *sp)
-{
-	unsigned long *stack;
-	int i;
-
-	stack = sp;
-	if (!stack) {
-		if (!task)
-			stack = (unsigned long *)current_stack_pointer();
-		else
-			stack = (unsigned long *)task->thread.ksp;
-	}
-	printk(KERN_DEFAULT "Stack:\n");
-	for (i = 0; i < 20; i++) {
-		if (((addr_t) stack & (THREAD_SIZE-1)) == 0)
-			break;
-		if (i % 4 == 0)
-			printk(KERN_DEFAULT "       ");
-		pr_cont("%016lx%c", *stack++, i % 4 == 3 ? '\n' : ' ');
-	}
-	show_trace(task, (unsigned long)sp);
-}
-
 static void show_last_breaking_event(struct pt_regs *regs)
 {
 	printk("Last Breaking-Event-Address:\n");
@@ -169,7 +148,7 @@ void show_regs(struct pt_regs *regs)
 	show_registers(regs);
 	/* Show stack backtrace if pt_regs is from kernel mode */
 	if (!user_mode(regs))
-		show_trace(NULL, regs->gprs[15]);
+		show_stack(NULL, (unsigned long *) regs->gprs[15]);
 	show_last_breaking_event(regs);
 }
 
-- 
cgit 


From ead1dec8ed0497c4a3f7fc36135899e5f26563ab Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Tue, 6 Jun 2017 13:55:42 +0200
Subject: s390/ipl: revert Load Normal semantics for LPAR CCW-type re-IPL

This reverts the two commits

7afbeb6df2aa ("s390/ipl: always use load normal for CCW-type re-IPL")
0f7451ff3ab8 ("s390/ipl: use load normal for LPAR re-ipl")

The two commits did not take into account that behavior of standby
memory changes fundamentally if the re-IPL method is changed from
Load Clear to Load Normal.

In case of the old re-IPL clear method all memory that was initially
in standby state will be put into standby state again within the
re-IPL process. Or in other words: memory that was brought online
before a re-IPL will be offline again after a reboot.

Given that we use different re-IPL methods depending on the hypervisor
and CCW-type vs SCSI re-IPL it is not easy to tell in advance when and
why memory will stay online or will be offline after a re-IPL.
This does also have other side effects, since memory that is online
from the beginning will be in ZONE_NORMAL by default vs ZONE_MOVABLE
for memory that is offline.

Therefore, before the change, a user could online and offline memory
easily since standby memory was always in ZONE_NORMAL.  After the
change, and a re-IPL, this depended on which memory parts were online
before the re-IPL.

From a usability point of view the current behavior is more than
suboptimal. Therefore revert these changes until we have a better
solution and get back to a consistent behavior. The bad thing about
this is that the time required for a re-IPL will be significantly
increased for configurations with several 100GB or 1TB of memory.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/ipl.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c
index e545ffe5155a..8e622bb52f7a 100644
--- a/arch/s390/kernel/ipl.c
+++ b/arch/s390/kernel/ipl.c
@@ -564,8 +564,6 @@ static struct kset *ipl_kset;
 
 static void __ipl_run(void *unused)
 {
-	if (MACHINE_IS_LPAR && ipl_info.type == IPL_TYPE_CCW)
-		diag308(DIAG308_LOAD_NORMAL_DUMP, NULL);
 	diag308(DIAG308_LOAD_CLEAR, NULL);
 	if (MACHINE_IS_VM)
 		__cpcmd("IPL", NULL, 0, NULL);
@@ -1088,10 +1086,7 @@ static void __reipl_run(void *unused)
 		break;
 	case REIPL_METHOD_CCW_DIAG:
 		diag308(DIAG308_SET, reipl_block_ccw);
-		if (MACHINE_IS_LPAR)
-			diag308(DIAG308_LOAD_NORMAL_DUMP, NULL);
-		else
-			diag308(DIAG308_LOAD_CLEAR, NULL);
+		diag308(DIAG308_LOAD_CLEAR, NULL);
 		break;
 	case REIPL_METHOD_FCP_RW_DIAG:
 		diag308(DIAG308_SET, reipl_block_fcp);
-- 
cgit 


From 60c497014e34af5aa0be56d0869c67fa2b5c3786 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Thu, 1 Jun 2017 11:04:04 +0200
Subject: s390/mm: use correct address space when enabling DAT

Right now the kernel uses the primary address space until finally the
switch to the correct home address space will be done when the idle
PSW will be loaded within psw_idle().

Correct this and simply use the home address space when DAT is enabled
for the first time.

This doesn't really fix a bug, but fixes odd behavior.

Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/mm/init.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

(limited to 'arch/s390')

diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c
index ee6a1d3d4983..0352f9f88c73 100644
--- a/arch/s390/mm/init.c
+++ b/arch/s390/mm/init.c
@@ -81,6 +81,7 @@ void __init paging_init(void)
 {
 	unsigned long max_zone_pfns[MAX_NR_ZONES];
 	unsigned long pgd_type, asce_bits;
+	psw_t psw;
 
 	init_mm.pgd = swapper_pg_dir;
 	if (VMALLOC_END > (1UL << 42)) {
@@ -100,7 +101,10 @@ void __init paging_init(void)
 	__ctl_load(S390_lowcore.kernel_asce, 1, 1);
 	__ctl_load(S390_lowcore.kernel_asce, 7, 7);
 	__ctl_load(S390_lowcore.kernel_asce, 13, 13);
-	__arch_local_irq_stosm(0x04);
+	psw.mask = __extract_psw();
+	psw_bits(psw).t = 1;
+	psw_bits(psw).as = PSW_AS_HOME;
+	__load_psw_mask(psw.mask);
 
 	sparse_memory_present_with_active_regions(MAX_NUMNODES);
 	sparse_init();
-- 
cgit 


From 8bb3fdd6863c3b6b84bbab750d6b35e889c1399d Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Sat, 3 Jun 2017 10:19:55 +0200
Subject: s390: rename psw_bits enums

The address space enums that must be used when modifying the address
space part of a psw with the psw_bits() macro can easily be confused
with the psw defines that are used to mask and compare directly the
mask part of a psw.
We have e.g. PSW_AS_PRIMARY vs PSW_ASC_PRIMARY.

To avoid confusion rename the PSW_AS_* enums to PSW_BITS_AS_*.

In addition also rename the PSW_AMODE_* enums, so they also follow the
same naming scheme: PSW_BITS_AMODE_*.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/include/asm/ptrace.h | 14 +++++++-------
 arch/s390/kernel/uprobes.c     |  8 ++++----
 arch/s390/kvm/gaccess.c        | 12 ++++++------
 arch/s390/kvm/gaccess.h        |  4 ++--
 arch/s390/kvm/guestdbg.c       |  6 +++---
 arch/s390/kvm/priv.c           |  6 +++---
 arch/s390/mm/init.c            |  2 +-
 7 files changed, 26 insertions(+), 26 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/ptrace.h b/arch/s390/include/asm/ptrace.h
index 99bc456cc26a..c8d13bcc9f5d 100644
--- a/arch/s390/include/asm/ptrace.h
+++ b/arch/s390/include/asm/ptrace.h
@@ -46,16 +46,16 @@ struct psw_bits {
 };
 
 enum {
-	PSW_AMODE_24BIT = 0,
-	PSW_AMODE_31BIT = 1,
-	PSW_AMODE_64BIT = 3
+	PSW_BITS_AMODE_24BIT = 0,
+	PSW_BITS_AMODE_31BIT = 1,
+	PSW_BITS_AMODE_64BIT = 3
 };
 
 enum {
-	PSW_AS_PRIMARY	 = 0,
-	PSW_AS_ACCREG	 = 1,
-	PSW_AS_SECONDARY = 2,
-	PSW_AS_HOME	 = 3
+	PSW_BITS_AS_PRIMARY	= 0,
+	PSW_BITS_AS_ACCREG	= 1,
+	PSW_BITS_AS_SECONDARY	= 2,
+	PSW_BITS_AS_HOME	= 3
 };
 
 #define psw_bits(__psw) (*({			\
diff --git a/arch/s390/kernel/uprobes.c b/arch/s390/kernel/uprobes.c
index 314e0ee3016a..0eec45b4575b 100644
--- a/arch/s390/kernel/uprobes.c
+++ b/arch/s390/kernel/uprobes.c
@@ -27,9 +27,9 @@ int arch_uprobe_analyze_insn(struct arch_uprobe *auprobe, struct mm_struct *mm,
 
 int arch_uprobe_pre_xol(struct arch_uprobe *auprobe, struct pt_regs *regs)
 {
-	if (psw_bits(regs->psw).eaba == PSW_AMODE_24BIT)
+	if (psw_bits(regs->psw).eaba == PSW_BITS_AMODE_24BIT)
 		return -EINVAL;
-	if (!is_compat_task() && psw_bits(regs->psw).eaba == PSW_AMODE_31BIT)
+	if (!is_compat_task() && psw_bits(regs->psw).eaba == PSW_BITS_AMODE_31BIT)
 		return -EINVAL;
 	clear_pt_regs_flag(regs, PIF_PER_TRAP);
 	auprobe->saved_per = psw_bits(regs->psw).r;
@@ -372,8 +372,8 @@ static void handle_insn_ril(struct arch_uprobe *auprobe, struct pt_regs *regs)
 
 bool arch_uprobe_skip_sstep(struct arch_uprobe *auprobe, struct pt_regs *regs)
 {
-	if ((psw_bits(regs->psw).eaba == PSW_AMODE_24BIT) ||
-	    ((psw_bits(regs->psw).eaba == PSW_AMODE_31BIT) &&
+	if ((psw_bits(regs->psw).eaba == PSW_BITS_AMODE_24BIT) ||
+	    ((psw_bits(regs->psw).eaba == PSW_BITS_AMODE_31BIT) &&
 	     !is_compat_task())) {
 		regs->psw.addr = __rewind_psw(regs->psw, UPROBE_SWBP_INSN_SIZE);
 		do_report_trap(regs, SIGILL, ILL_ILLADR, NULL);
diff --git a/arch/s390/kvm/gaccess.c b/arch/s390/kvm/gaccess.c
index 9da243d94cc3..6ad4a9797de8 100644
--- a/arch/s390/kvm/gaccess.c
+++ b/arch/s390/kvm/gaccess.c
@@ -557,20 +557,20 @@ static int get_vcpu_asce(struct kvm_vcpu *vcpu, union asce *asce,
 		return 0;
 	}
 
-	if (mode == GACC_IFETCH)
-		psw.as = psw.as == PSW_AS_HOME ? PSW_AS_HOME : PSW_AS_PRIMARY;
+	if ((mode == GACC_IFETCH) && (psw.as != PSW_BITS_AS_HOME))
+		psw.as = PSW_BITS_AS_PRIMARY;
 
 	switch (psw.as) {
-	case PSW_AS_PRIMARY:
+	case PSW_BITS_AS_PRIMARY:
 		asce->val = vcpu->arch.sie_block->gcr[1];
 		return 0;
-	case PSW_AS_SECONDARY:
+	case PSW_BITS_AS_SECONDARY:
 		asce->val = vcpu->arch.sie_block->gcr[7];
 		return 0;
-	case PSW_AS_HOME:
+	case PSW_BITS_AS_HOME:
 		asce->val = vcpu->arch.sie_block->gcr[13];
 		return 0;
-	case PSW_AS_ACCREG:
+	case PSW_BITS_AS_ACCREG:
 		rc = ar_translation(vcpu, asce, ar, mode);
 		if (rc > 0)
 			return trans_exc(vcpu, rc, ga, ar, mode, PROT_TYPE_ALC);
diff --git a/arch/s390/kvm/gaccess.h b/arch/s390/kvm/gaccess.h
index 7ce47fd36f28..bec42b852246 100644
--- a/arch/s390/kvm/gaccess.h
+++ b/arch/s390/kvm/gaccess.h
@@ -57,9 +57,9 @@ static inline unsigned long kvm_s390_logical_to_effective(struct kvm_vcpu *vcpu,
 {
 	psw_t *psw = &vcpu->arch.sie_block->gpsw;
 
-	if (psw_bits(*psw).eaba == PSW_AMODE_64BIT)
+	if (psw_bits(*psw).eaba == PSW_BITS_AMODE_64BIT)
 		return ga;
-	if (psw_bits(*psw).eaba == PSW_AMODE_31BIT)
+	if (psw_bits(*psw).eaba == PSW_BITS_AMODE_31BIT)
 		return ga & ((1UL << 31) - 1);
 	return ga & ((1UL << 24) - 1);
 }
diff --git a/arch/s390/kvm/guestdbg.c b/arch/s390/kvm/guestdbg.c
index 23d9a4e12da1..c2e0ddc1356e 100644
--- a/arch/s390/kvm/guestdbg.c
+++ b/arch/s390/kvm/guestdbg.c
@@ -613,15 +613,15 @@ int kvm_s390_handle_per_event(struct kvm_vcpu *vcpu)
 		 * instruction. Check primary and home space-switch-event
 		 * controls. (theoretically home -> home produced no event)
 		 */
-		if (((new_as == PSW_AS_HOME) ^ old_as_is_home(vcpu)) &&
-		     (pssec(vcpu) || hssec(vcpu)))
+		if (((new_as == PSW_BITS_AS_HOME) ^ old_as_is_home(vcpu)) &&
+		    (pssec(vcpu) || hssec(vcpu)))
 			vcpu->arch.sie_block->iprcc = PGM_SPACE_SWITCH;
 
 		/*
 		 * PT, PTI, PR, PC instruction operate on primary AS only. Check
 		 * if the primary-space-switch-event control was or got set.
 		 */
-		if (new_as == PSW_AS_PRIMARY && !old_as_is_home(vcpu) &&
+		if (new_as == PSW_BITS_AS_PRIMARY && !old_as_is_home(vcpu) &&
 		    (pssec(vcpu) || old_ssec(vcpu)))
 			vcpu->arch.sie_block->iprcc = PGM_SPACE_SWITCH;
 	}
diff --git a/arch/s390/kvm/priv.c b/arch/s390/kvm/priv.c
index c03106c428cf..e9dd7efc57b8 100644
--- a/arch/s390/kvm/priv.c
+++ b/arch/s390/kvm/priv.c
@@ -361,7 +361,7 @@ static int handle_sske(struct kvm_vcpu *vcpu)
 		}
 	}
 	if (m3 & SSKE_MB) {
-		if (psw_bits(vcpu->arch.sie_block->gpsw).eaba == PSW_AMODE_64BIT)
+		if (psw_bits(vcpu->arch.sie_block->gpsw).eaba == PSW_BITS_AMODE_64BIT)
 			vcpu->run->s.regs.gprs[reg2] &= ~PAGE_MASK;
 		else
 			vcpu->run->s.regs.gprs[reg2] &= ~0xfffff000UL;
@@ -901,7 +901,7 @@ static int handle_pfmf(struct kvm_vcpu *vcpu)
 		/* only support 2G frame size if EDAT2 is available and we are
 		   not in 24-bit addressing mode */
 		if (!test_kvm_facility(vcpu->kvm, 78) ||
-		    psw_bits(vcpu->arch.sie_block->gpsw).eaba == PSW_AMODE_24BIT)
+		    psw_bits(vcpu->arch.sie_block->gpsw).eaba == PSW_BITS_AMODE_24BIT)
 			return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
 		end = (start + (1UL << 31)) & ~((1UL << 31) - 1);
 		break;
@@ -938,7 +938,7 @@ static int handle_pfmf(struct kvm_vcpu *vcpu)
 		start += PAGE_SIZE;
 	}
 	if (vcpu->run->s.regs.gprs[reg1] & PFMF_FSC) {
-		if (psw_bits(vcpu->arch.sie_block->gpsw).eaba == PSW_AMODE_64BIT) {
+		if (psw_bits(vcpu->arch.sie_block->gpsw).eaba == PSW_BITS_AMODE_64BIT) {
 			vcpu->run->s.regs.gprs[reg2] = end;
 		} else {
 			vcpu->run->s.regs.gprs[reg2] &= ~0xffffffffUL;
diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c
index 0352f9f88c73..bc8c301f82b6 100644
--- a/arch/s390/mm/init.c
+++ b/arch/s390/mm/init.c
@@ -103,7 +103,7 @@ void __init paging_init(void)
 	__ctl_load(S390_lowcore.kernel_asce, 13, 13);
 	psw.mask = __extract_psw();
 	psw_bits(psw).t = 1;
-	psw_bits(psw).as = PSW_AS_HOME;
+	psw_bits(psw).as = PSW_BITS_AS_HOME;
 	__load_psw_mask(psw.mask);
 
 	sparse_memory_present_with_active_regions(MAX_NUMNODES);
-- 
cgit 


From a752598254016d2f9b4415d43a6402fe083f70b2 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Sat, 3 Jun 2017 10:56:07 +0200
Subject: s390: rename struct psw_bits members

Rename a couple of the struct psw_bits members so it is more obvious
for what they are good. Initially I thought using the single character
names from the PoP would be sufficient and obvious, but admittedly
that is not true.

The current implementation is not easy to use, if one has to look into
the source file to figure out which member represents the 'per' bit
(which is the 'r' member).

Therefore rename the members to sane names that are identical to the
uapi psw mask defines:

r -> per
i -> io
e -> ext
t -> dat
m -> mcheck
w -> wait
p -> pstate

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/include/asm/ptrace.h  | 38 +++++++++++++++++++-------------------
 arch/s390/kernel/dumpstack.c    |  4 ++--
 arch/s390/kernel/perf_cpum_sf.c | 10 +++++-----
 arch/s390/kernel/uprobes.c      |  4 ++--
 arch/s390/kvm/gaccess.c         | 10 +++++-----
 arch/s390/kvm/priv.c            |  2 +-
 arch/s390/mm/init.c             |  2 +-
 7 files changed, 35 insertions(+), 35 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/ptrace.h b/arch/s390/include/asm/ptrace.h
index c8d13bcc9f5d..004f54909235 100644
--- a/arch/s390/include/asm/ptrace.h
+++ b/arch/s390/include/asm/ptrace.h
@@ -24,25 +24,25 @@
 			 PSW_MASK_PSTATE | PSW_ASC_PRIMARY)
 
 struct psw_bits {
-	unsigned long	   :  1;
-	unsigned long r	   :  1; /* PER-Mask */
-	unsigned long	   :  3;
-	unsigned long t	   :  1; /* DAT Mode */
-	unsigned long i	   :  1; /* Input/Output Mask */
-	unsigned long e	   :  1; /* External Mask */
-	unsigned long key  :  4; /* PSW Key */
-	unsigned long	   :  1;
-	unsigned long m	   :  1; /* Machine-Check Mask */
-	unsigned long w	   :  1; /* Wait State */
-	unsigned long p	   :  1; /* Problem State */
-	unsigned long as   :  2; /* Address Space Control */
-	unsigned long cc   :  2; /* Condition Code */
-	unsigned long pm   :  4; /* Program Mask */
-	unsigned long ri   :  1; /* Runtime Instrumentation */
-	unsigned long	   :  6;
-	unsigned long eaba :  2; /* Addressing Mode */
-	unsigned long	   : 31;
-	unsigned long ia   : 64; /* Instruction Address */
+	unsigned long	     :	1;
+	unsigned long per    :	1; /* PER-Mask */
+	unsigned long	     :	3;
+	unsigned long dat    :	1; /* DAT Mode */
+	unsigned long io     :	1; /* Input/Output Mask */
+	unsigned long ext    :	1; /* External Mask */
+	unsigned long key    :	4; /* PSW Key */
+	unsigned long	     :	1;
+	unsigned long mcheck :	1; /* Machine-Check Mask */
+	unsigned long wait   :	1; /* Wait State */
+	unsigned long pstate :	1; /* Problem State */
+	unsigned long as     :	2; /* Address Space Control */
+	unsigned long cc     :	2; /* Condition Code */
+	unsigned long pm     :	4; /* Program Mask */
+	unsigned long ri     :	1; /* Runtime Instrumentation */
+	unsigned long	     :	6;
+	unsigned long eaba   :	2; /* Addressing Mode */
+	unsigned long	     : 31;
+	unsigned long ia     : 64; /* Instruction Address */
 };
 
 enum {
diff --git a/arch/s390/kernel/dumpstack.c b/arch/s390/kernel/dumpstack.c
index 62904890d127..dab78babfab6 100644
--- a/arch/s390/kernel/dumpstack.c
+++ b/arch/s390/kernel/dumpstack.c
@@ -128,8 +128,8 @@ void show_registers(struct pt_regs *regs)
 		pr_cont(" (%pSR)", (void *)regs->psw.addr);
 	pr_cont("\n");
 	printk("           R:%x T:%x IO:%x EX:%x Key:%x M:%x W:%x "
-	       "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);
+	       "P:%x AS:%x CC:%x PM:%x", psw->per, psw->dat, psw->io, psw->ext,
+	       psw->key, psw->mcheck, psw->wait, psw->pstate, psw->as, psw->cc, psw->pm);
 	pr_cont(" RI:%x EA:%x\n", psw->ri, psw->eaba);
 	printk("%s GPRS: %016lx %016lx %016lx %016lx\n", mode,
 	       regs->gprs[0], regs->gprs[1], regs->gprs[2], regs->gprs[3]);
diff --git a/arch/s390/kernel/perf_cpum_sf.c b/arch/s390/kernel/perf_cpum_sf.c
index ca960d0370d5..0c82f7903fc7 100644
--- a/arch/s390/kernel/perf_cpum_sf.c
+++ b/arch/s390/kernel/perf_cpum_sf.c
@@ -995,11 +995,11 @@ static int perf_push_sample(struct perf_event *event, struct sf_raw_sample *sfr)
 	regs.int_parm = CPU_MF_INT_SF_PRA;
 	sde_regs = (struct perf_sf_sde_regs *) &regs.int_parm_long;
 
-	psw_bits(regs.psw).ia = sfr->basic.ia;
-	psw_bits(regs.psw).t  = sfr->basic.T;
-	psw_bits(regs.psw).w  = sfr->basic.W;
-	psw_bits(regs.psw).p  = sfr->basic.P;
-	psw_bits(regs.psw).as = sfr->basic.AS;
+	psw_bits(regs.psw).ia	= sfr->basic.ia;
+	psw_bits(regs.psw).dat	= sfr->basic.T;
+	psw_bits(regs.psw).wait = sfr->basic.W;
+	psw_bits(regs.psw).per	= sfr->basic.P;
+	psw_bits(regs.psw).as	= sfr->basic.AS;
 
 	/*
 	 * Use the hardware provided configuration level to decide if the
diff --git a/arch/s390/kernel/uprobes.c b/arch/s390/kernel/uprobes.c
index 0eec45b4575b..d94baa8db507 100644
--- a/arch/s390/kernel/uprobes.c
+++ b/arch/s390/kernel/uprobes.c
@@ -32,7 +32,7 @@ int arch_uprobe_pre_xol(struct arch_uprobe *auprobe, struct pt_regs *regs)
 	if (!is_compat_task() && psw_bits(regs->psw).eaba == PSW_BITS_AMODE_31BIT)
 		return -EINVAL;
 	clear_pt_regs_flag(regs, PIF_PER_TRAP);
-	auprobe->saved_per = psw_bits(regs->psw).r;
+	auprobe->saved_per = psw_bits(regs->psw).per;
 	auprobe->saved_int_code = regs->int_code;
 	regs->int_code = UPROBE_TRAP_NR;
 	regs->psw.addr = current->utask->xol_vaddr;
@@ -81,7 +81,7 @@ int arch_uprobe_post_xol(struct arch_uprobe *auprobe, struct pt_regs *regs)
 
 	clear_tsk_thread_flag(current, TIF_UPROBE_SINGLESTEP);
 	update_cr_regs(current);
-	psw_bits(regs->psw).r = auprobe->saved_per;
+	psw_bits(regs->psw).per = auprobe->saved_per;
 	regs->int_code = auprobe->saved_int_code;
 
 	if (fixup & FIXUP_PSW_NORMAL)
diff --git a/arch/s390/kvm/gaccess.c b/arch/s390/kvm/gaccess.c
index 6ad4a9797de8..e0f7d5fc7efd 100644
--- a/arch/s390/kvm/gaccess.c
+++ b/arch/s390/kvm/gaccess.c
@@ -551,7 +551,7 @@ static int get_vcpu_asce(struct kvm_vcpu *vcpu, union asce *asce,
 	int rc;
 	struct psw_bits psw = psw_bits(vcpu->arch.sie_block->gpsw);
 
-	if (!psw.t) {
+	if (!psw.dat) {
 		asce->val = 0;
 		asce->r = 1;
 		return 0;
@@ -771,7 +771,7 @@ static int low_address_protection_enabled(struct kvm_vcpu *vcpu,
 
 	if (!ctlreg0.lap)
 		return 0;
-	if (psw_bits(*psw).t && asce.p)
+	if (psw_bits(*psw).dat && asce.p)
 		return 0;
 	return 1;
 }
@@ -790,7 +790,7 @@ static int guest_page_range(struct kvm_vcpu *vcpu, unsigned long ga, u8 ar,
 			return trans_exc(vcpu, PGM_PROTECTION, ga, ar, mode,
 					 PROT_TYPE_LA);
 		ga &= PAGE_MASK;
-		if (psw_bits(*psw).t) {
+		if (psw_bits(*psw).dat) {
 			rc = guest_translate(vcpu, ga, pages, asce, mode);
 			if (rc < 0)
 				return rc;
@@ -831,7 +831,7 @@ int access_guest(struct kvm_vcpu *vcpu, unsigned long ga, u8 ar, void *data,
 		pages = vmalloc(nr_pages * sizeof(unsigned long));
 	if (!pages)
 		return -ENOMEM;
-	need_ipte_lock = psw_bits(*psw).t && !asce.r;
+	need_ipte_lock = psw_bits(*psw).dat && !asce.r;
 	if (need_ipte_lock)
 		ipte_lock(vcpu);
 	rc = guest_page_range(vcpu, ga, ar, pages, nr_pages, asce, mode);
@@ -899,7 +899,7 @@ int guest_translate_address(struct kvm_vcpu *vcpu, unsigned long gva, u8 ar,
 					 mode, PROT_TYPE_LA);
 	}
 
-	if (psw_bits(*psw).t && !asce.r) {	/* Use DAT? */
+	if (psw_bits(*psw).dat && !asce.r) {	/* Use DAT? */
 		rc = guest_translate(vcpu, gva, gpa, asce, mode);
 		if (rc > 0)
 			return trans_exc(vcpu, rc, gva, 0, mode, PROT_TYPE_DAT);
diff --git a/arch/s390/kvm/priv.c b/arch/s390/kvm/priv.c
index e9dd7efc57b8..e53292a89257 100644
--- a/arch/s390/kvm/priv.c
+++ b/arch/s390/kvm/priv.c
@@ -374,7 +374,7 @@ static int handle_sske(struct kvm_vcpu *vcpu)
 static int handle_ipte_interlock(struct kvm_vcpu *vcpu)
 {
 	vcpu->stat.instruction_ipte_interlock++;
-	if (psw_bits(vcpu->arch.sie_block->gpsw).p)
+	if (psw_bits(vcpu->arch.sie_block->gpsw).pstate)
 		return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);
 	wait_event(vcpu->kvm->arch.ipte_wq, !ipte_lock_held(vcpu));
 	kvm_s390_retry_instr(vcpu);
diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c
index bc8c301f82b6..3348e60dd8ad 100644
--- a/arch/s390/mm/init.c
+++ b/arch/s390/mm/init.c
@@ -102,7 +102,7 @@ void __init paging_init(void)
 	__ctl_load(S390_lowcore.kernel_asce, 7, 7);
 	__ctl_load(S390_lowcore.kernel_asce, 13, 13);
 	psw.mask = __extract_psw();
-	psw_bits(psw).t = 1;
+	psw_bits(psw).dat = 1;
 	psw_bits(psw).as = PSW_BITS_AS_HOME;
 	__load_psw_mask(psw.mask);
 
-- 
cgit 


From 23fefe119ceb5fb0c7d3321010620010a4eddb18 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Wed, 7 Jun 2017 14:10:24 +0200
Subject: s390/kvm: avoid global config of vm.alloc_pgste=1

The system control vm.alloc_pgste is used to control the size of the
page tables, either 2K or 4K. The idea is that a KVM host sets the
vm.alloc_pgste control to 1 which causes *all* new processes to run
with 4K page tables. For a non-kvm system the control should stay off
to save on memory used for page tables.

Trouble is that distributions choose to set the control globally to
be able to run KVM guests. This wastes memory on non-KVM systems.

Introduce the PT_S390_PGSTE ELF segment type to "mark" the qemu
executable with it. All executables with this (empty) segment in
its ELF phdr array will be started with 4K page tables. Any executable
without PT_S390_PGSTE will run with the default 2K page tables.

This removes the need to set vm.alloc_pgste=1 for a KVM host and
minimizes the waste of memory for page tables.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/Kconfig                   |  1 +
 arch/s390/include/asm/elf.h         | 32 ++++++++++++++++++++++++++++++++
 arch/s390/include/asm/mmu_context.h |  4 +++-
 arch/s390/include/asm/ptrace.h      |  2 ++
 arch/s390/include/asm/thread_info.h |  1 +
 arch/s390/kernel/entry.S            | 15 ++++++++++++++-
 6 files changed, 53 insertions(+), 2 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
index e7ff58150e8f..bb11f9f30c8d 100644
--- a/arch/s390/Kconfig
+++ b/arch/s390/Kconfig
@@ -64,6 +64,7 @@ config ARCH_SUPPORTS_UPROBES
 
 config S390
 	def_bool y
+	select ARCH_BINFMT_ELF_STATE
 	select ARCH_HAS_DEVMEM_IS_ALLOWED
 	select ARCH_HAS_ELF_RANDOMIZE
 	select ARCH_HAS_GCOV_PROFILE_ALL
diff --git a/arch/s390/include/asm/elf.h b/arch/s390/include/asm/elf.h
index e8f623041769..ec024c08dabe 100644
--- a/arch/s390/include/asm/elf.h
+++ b/arch/s390/include/asm/elf.h
@@ -117,6 +117,9 @@
 #define ELF_DATA	ELFDATA2MSB
 #define ELF_ARCH	EM_S390
 
+/* s390 specific phdr types */
+#define PT_S390_PGSTE	0x70000000
+
 /*
  * ELF register definitions..
  */
@@ -151,6 +154,35 @@ extern unsigned int vdso_enabled;
 	 && (x)->e_ident[EI_CLASS] == ELF_CLASS)
 #define compat_start_thread	start_thread31
 
+struct arch_elf_state {
+	int rc;
+};
+
+#define INIT_ARCH_ELF_STATE { .rc = 0 }
+
+#define arch_check_elf(ehdr, interp, interp_ehdr, state) (0)
+#ifdef CONFIG_PGSTE
+#define arch_elf_pt_proc(ehdr, phdr, elf, interp, state)	\
+({								\
+	struct arch_elf_state *_state = state;			\
+	if ((phdr)->p_type == PT_S390_PGSTE &&			\
+	    !page_table_allocate_pgste &&			\
+	    !test_thread_flag(TIF_PGSTE) &&			\
+	    !current->mm->context.alloc_pgste) {		\
+		set_thread_flag(TIF_PGSTE);			\
+		set_pt_regs_flag(task_pt_regs(current),		\
+				 PIF_SYSCALL_RESTART);		\
+		_state->rc = -EAGAIN;				\
+	}							\
+	_state->rc;						\
+})
+#else
+#define arch_elf_pt_proc(ehdr, phdr, elf, interp, state)	\
+({								\
+	(state)->rc;						\
+})
+#endif
+
 /* For SVR4/S390 the function pointer to be registered with `atexit` is
    passed in R14. */
 #define ELF_PLAT_INIT(_r, load_addr) \
diff --git a/arch/s390/include/asm/mmu_context.h b/arch/s390/include/asm/mmu_context.h
index 8712e11bead4..4541ac44b35f 100644
--- a/arch/s390/include/asm/mmu_context.h
+++ b/arch/s390/include/asm/mmu_context.h
@@ -25,7 +25,9 @@ static inline int init_new_context(struct task_struct *tsk,
 	mm->context.gmap_asce = 0;
 	mm->context.flush_mm = 0;
 #ifdef CONFIG_PGSTE
-	mm->context.alloc_pgste = page_table_allocate_pgste;
+	mm->context.alloc_pgste = page_table_allocate_pgste ||
+		test_thread_flag(TIF_PGSTE) ||
+		current->mm->context.alloc_pgste;
 	mm->context.has_pgste = 0;
 	mm->context.use_skey = 0;
 	mm->context.use_cmma = 0;
diff --git a/arch/s390/include/asm/ptrace.h b/arch/s390/include/asm/ptrace.h
index 004f54909235..853b01245c20 100644
--- a/arch/s390/include/asm/ptrace.h
+++ b/arch/s390/include/asm/ptrace.h
@@ -11,9 +11,11 @@
 
 #define PIF_SYSCALL		0	/* inside a system call */
 #define PIF_PER_TRAP		1	/* deliver sigtrap on return to user */
+#define PIF_SYSCALL_RESTART	2	/* restart the current system call */
 
 #define _PIF_SYSCALL		_BITUL(PIF_SYSCALL)
 #define _PIF_PER_TRAP		_BITUL(PIF_PER_TRAP)
+#define _PIF_SYSCALL_RESTART	_BITUL(PIF_SYSCALL_RESTART)
 
 #ifndef __ASSEMBLY__
 
diff --git a/arch/s390/include/asm/thread_info.h b/arch/s390/include/asm/thread_info.h
index 0b3ee083a665..1aecf432c48d 100644
--- a/arch/s390/include/asm/thread_info.h
+++ b/arch/s390/include/asm/thread_info.h
@@ -58,6 +58,7 @@ int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src);
 #define TIF_UPROBE		3	/* breakpointed or single-stepping */
 #define TIF_GUARDED_STORAGE	4	/* load guarded storage control block */
 #define TIF_PATCH_PENDING	5	/* pending live patching update */
+#define TIF_PGSTE		6	/* New mm's will use 4K page tables */
 
 #define TIF_31BIT		16	/* 32bit process */
 #define TIF_MEMDIE		17	/* is terminating due to OOM killer */
diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S
index 6315037335ba..0c0138c7dfc7 100644
--- a/arch/s390/kernel/entry.S
+++ b/arch/s390/kernel/entry.S
@@ -52,7 +52,7 @@ _TIF_TRACE	= (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | _TIF_SECCOMP | \
 		   _TIF_SYSCALL_TRACEPOINT)
 _CIF_WORK	= (_CIF_MCCK_PENDING | _CIF_ASCE_PRIMARY | \
 		   _CIF_ASCE_SECONDARY | _CIF_FPU)
-_PIF_WORK	= (_PIF_PER_TRAP)
+_PIF_WORK	= (_PIF_PER_TRAP | _PIF_SYSCALL_RESTART)
 
 #define BASED(name) name-cleanup_critical(%r13)
 
@@ -334,6 +334,8 @@ ENTRY(system_call)
 	jo	.Lsysc_mcck_pending
 	TSTMSK	__TI_flags(%r12),_TIF_NEED_RESCHED
 	jo	.Lsysc_reschedule
+	TSTMSK	__PT_FLAGS(%r11),_PIF_SYSCALL_RESTART
+	jo	.Lsysc_syscall_restart
 #ifdef CONFIG_UPROBES
 	TSTMSK	__TI_flags(%r12),_TIF_UPROBE
 	jo	.Lsysc_uprobe_notify
@@ -347,6 +349,8 @@ ENTRY(system_call)
 	jo	.Lsysc_patch_pending	# handle live patching just before
 					# signals and possible syscall restart
 #endif
+	TSTMSK	__PT_FLAGS(%r11),_PIF_SYSCALL_RESTART
+	jo	.Lsysc_syscall_restart
 	TSTMSK	__TI_flags(%r12),_TIF_SIGPENDING
 	jo	.Lsysc_sigpending
 	TSTMSK	__TI_flags(%r12),_TIF_NOTIFY_RESUME
@@ -447,6 +451,15 @@ ENTRY(system_call)
 	larl	%r14,.Lsysc_return
 	jg	do_per_trap
 
+#
+# _PIF_SYSCALL_RESTART is set, repeat the current system call
+#
+.Lsysc_syscall_restart:
+	ni	__PT_FLAGS+7(%r11),255-_PIF_SYSCALL_RESTART
+	lmg	%r1,%r7,__PT_R1(%r11)	# load svc arguments
+	lg	%r2,__PT_ORIG_GPR2(%r11)
+	j	.Lsysc_do_svc
+
 #
 # call tracehook_report_syscall_entry/tracehook_report_syscall_exit before
 # and after the system call
-- 
cgit 


From f044f4c588985867d17499e51460053a96f9ff7a Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Mon, 12 Jun 2017 08:52:56 +0200
Subject: s390/fpu: export save_fpu_regs for all configs

The save_fpu_regs function is a general API that is supposed to be
usable for modules as well. Remove the #ifdef that hides the symbol
for CONFIG_KVM=n.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/entry.S | 2 --
 1 file changed, 2 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S
index 0c0138c7dfc7..e83aff630bcf 100644
--- a/arch/s390/kernel/entry.S
+++ b/arch/s390/kernel/entry.S
@@ -894,9 +894,7 @@ ENTRY(save_fpu_regs)
 	oi	__LC_CPU_FLAGS+7,_CIF_FPU
 	br	%r14
 .Lsave_fpu_regs_end:
-#if IS_ENABLED(CONFIG_KVM)
 EXPORT_SYMBOL(save_fpu_regs)
-#endif
 
 /*
  * Load floating-point controls and floating-point or vector registers.
-- 
cgit 


From c929500d7a5aaea4f2eeba10816bc5341c66ae57 Mon Sep 17 00:00:00 2001
From: QingFeng Hao <haoqf@linux.vnet.ibm.com>
Date: Wed, 7 Jun 2017 11:30:42 +0200
Subject: s390/nmi: s390: New low level handling for machine check happening in
 guest

Add the logic to check if the machine check happens when the guest is
running. If yes, set the exit reason -EINTR in the machine check's
interrupt handler. Refactor s390_do_machine_check to avoid panicing
the host for some kinds of machine checks which happen
when guest is running.
Reinject the instruction processing damage's machine checks including
Delayed Access Exception instead of damaging the host if it happens
in the guest because it could be caused by improper update on TLB entry
or other software case and impacts the guest only.

Signed-off-by: QingFeng Hao <haoqf@linux.vnet.ibm.com>
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
 arch/s390/include/asm/nmi.h       |  7 ++++++
 arch/s390/include/asm/processor.h |  2 ++
 arch/s390/kernel/asm-offsets.c    |  3 +++
 arch/s390/kernel/entry.S          | 13 +++++++++-
 arch/s390/kernel/nmi.c            | 50 +++++++++++++++++++++++++++++++--------
 5 files changed, 64 insertions(+), 11 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/nmi.h b/arch/s390/include/asm/nmi.h
index e3e8895f5d3e..13623b9991d4 100644
--- a/arch/s390/include/asm/nmi.h
+++ b/arch/s390/include/asm/nmi.h
@@ -14,7 +14,14 @@
 #include <linux/const.h>
 #include <linux/types.h>
 
+#define MCIC_SUBCLASS_MASK	(1ULL<<63 | 1ULL<<62 | 1ULL<<61 | \
+				1ULL<<59 | 1ULL<<58 | 1ULL<<56 | \
+				1ULL<<55 | 1ULL<<54 | 1ULL<<53 | \
+				1ULL<<52 | 1ULL<<47 | 1ULL<<46 | \
+				1ULL<<45 | 1ULL<<44)
 #define MCCK_CODE_SYSTEM_DAMAGE		_BITUL(63)
+#define MCCK_CODE_EXT_DAMAGE		_BITUL(63 - 5)
+#define MCCK_CODE_CP			_BITUL(63 - 9)
 #define MCCK_CODE_CPU_TIMER_VALID	_BITUL(63 - 46)
 #define MCCK_CODE_PSW_MWP_VALID		_BITUL(63 - 20)
 #define MCCK_CODE_PSW_IA_VALID		_BITUL(63 - 23)
diff --git a/arch/s390/include/asm/processor.h b/arch/s390/include/asm/processor.h
index 60d395fdc864..5b1b247dfbca 100644
--- a/arch/s390/include/asm/processor.h
+++ b/arch/s390/include/asm/processor.h
@@ -20,6 +20,7 @@
 #define CIF_FPU			4	/* restore FPU registers */
 #define CIF_IGNORE_IRQ		5	/* ignore interrupt (for udelay) */
 #define CIF_ENABLED_WAIT	6	/* in enabled wait state */
+#define CIF_MCCK_GUEST		7	/* machine check happening in guest */
 
 #define _CIF_MCCK_PENDING	_BITUL(CIF_MCCK_PENDING)
 #define _CIF_ASCE_PRIMARY	_BITUL(CIF_ASCE_PRIMARY)
@@ -28,6 +29,7 @@
 #define _CIF_FPU		_BITUL(CIF_FPU)
 #define _CIF_IGNORE_IRQ		_BITUL(CIF_IGNORE_IRQ)
 #define _CIF_ENABLED_WAIT	_BITUL(CIF_ENABLED_WAIT)
+#define _CIF_MCCK_GUEST		_BITUL(CIF_MCCK_GUEST)
 
 #ifndef __ASSEMBLY__
 
diff --git a/arch/s390/kernel/asm-offsets.c b/arch/s390/kernel/asm-offsets.c
index 6bb29633e1f1..b65c414b6c0e 100644
--- a/arch/s390/kernel/asm-offsets.c
+++ b/arch/s390/kernel/asm-offsets.c
@@ -58,6 +58,9 @@ int main(void)
 	OFFSET(__SF_BACKCHAIN, stack_frame, back_chain);
 	OFFSET(__SF_GPRS, stack_frame, gprs);
 	OFFSET(__SF_EMPTY, stack_frame, empty1);
+	OFFSET(__SF_SIE_CONTROL, stack_frame, empty1[0]);
+	OFFSET(__SF_SIE_SAVEAREA, stack_frame, empty1[1]);
+	OFFSET(__SF_SIE_REASON, stack_frame, empty1[2]);
 	BLANK();
 	/* timeval/timezone offsets for use by vdso */
 	OFFSET(__VDSO_UPD_COUNT, vdso_data, tb_update_count);
diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S
index a5f5d3bb3dbc..9b48196ebf40 100644
--- a/arch/s390/kernel/entry.S
+++ b/arch/s390/kernel/entry.S
@@ -225,6 +225,7 @@ ENTRY(sie64a)
 	jnz	.Lsie_skip
 	TSTMSK	__LC_CPU_FLAGS,_CIF_FPU
 	jo	.Lsie_skip			# exit if fp/vx regs changed
+.Lsie_entry:
 	sie	0(%r14)
 .Lsie_skip:
 	ni	__SIE_PROG0C+3(%r14),0xfe	# no longer in SIE
@@ -1102,7 +1103,13 @@ cleanup_critical:
 	.quad	.Lsie_done
 
 .Lcleanup_sie:
-	lg	%r9,__SF_EMPTY(%r15)		# get control block pointer
+	cghi    %r11,__LC_SAVE_AREA_ASYNC 	#Is this in normal interrupt?
+	je      1f
+	slg     %r9,BASED(.Lsie_crit_mcck_start)
+	clg     %r9,BASED(.Lsie_crit_mcck_length)
+	jh      1f
+	oi      __LC_CPU_FLAGS+7, _CIF_MCCK_GUEST
+1:	lg	%r9,__SF_EMPTY(%r15)		# get control block pointer
 	ni	__SIE_PROG0C+3(%r9),0xfe	# no longer in SIE
 	lctlg	%c1,%c1,__LC_USER_ASCE		# load primary asce
 	larl	%r9,sie_exit			# skip forward to sie_exit
@@ -1274,6 +1281,10 @@ cleanup_critical:
 	.quad	.Lsie_gmap
 .Lsie_critical_length:
 	.quad	.Lsie_done - .Lsie_gmap
+.Lsie_crit_mcck_start:
+	.quad   .Lsie_entry
+.Lsie_crit_mcck_length:
+	.quad   .Lsie_skip - .Lsie_entry
 #endif
 
 	.section .rodata, "a"
diff --git a/arch/s390/kernel/nmi.c b/arch/s390/kernel/nmi.c
index 985589523970..958cc3352faa 100644
--- a/arch/s390/kernel/nmi.c
+++ b/arch/s390/kernel/nmi.c
@@ -25,6 +25,7 @@
 #include <asm/crw.h>
 #include <asm/switch_to.h>
 #include <asm/ctl_reg.h>
+#include <asm/asm-offsets.h>
 
 struct mcck_struct {
 	unsigned int kill_task : 1;
@@ -280,6 +281,8 @@ static int notrace s390_validate_registers(union mci mci, int umode)
 #define ED_STP_ISLAND	6	/* External damage STP island check */
 #define ED_STP_SYNC	7	/* External damage STP sync check */
 
+#define MCCK_CODE_NO_GUEST	(MCCK_CODE_CP | MCCK_CODE_EXT_DAMAGE)
+
 /*
  * machine check handler.
  */
@@ -291,6 +294,7 @@ void notrace s390_do_machine_check(struct pt_regs *regs)
 	struct mcck_struct *mcck;
 	unsigned long long tmp;
 	union mci mci;
+	unsigned long mcck_dam_code;
 
 	nmi_enter();
 	inc_irq_stat(NMI_NMI);
@@ -301,7 +305,13 @@ void notrace s390_do_machine_check(struct pt_regs *regs)
 		/* System damage -> stopping machine */
 		s390_handle_damage();
 	}
-	if (mci.pd) {
+
+	/*
+	 * Reinject the instruction processing damages' machine checks
+	 * including Delayed Access Exception into the guest
+	 * instead of damaging the host if they happen in the guest.
+	 */
+	if (mci.pd && !test_cpu_flag(CIF_MCCK_GUEST)) {
 		if (mci.b) {
 			/* Processing backup -> verify if we can survive this */
 			u64 z_mcic, o_mcic, t_mcic;
@@ -358,15 +368,22 @@ void notrace s390_do_machine_check(struct pt_regs *regs)
 		if (mcck->stp_queue)
 			set_cpu_flag(CIF_MCCK_PENDING);
 	}
-	if (mci.se)
-		/* Storage error uncorrected */
-		s390_handle_damage();
-	if (mci.ke)
-		/* Storage key-error uncorrected */
-		s390_handle_damage();
-	if (mci.ds && mci.fa)
-		/* Storage degradation */
-		s390_handle_damage();
+
+	/*
+	 * Reinject storage related machine checks into the guest if they
+	 * happen when the guest is running.
+	 */
+	if (!test_cpu_flag(CIF_MCCK_GUEST)) {
+		if (mci.se)
+			/* Storage error uncorrected */
+			s390_handle_damage();
+		if (mci.ke)
+			/* Storage key-error uncorrected */
+			s390_handle_damage();
+		if (mci.ds && mci.fa)
+			/* Storage degradation */
+			s390_handle_damage();
+	}
 	if (mci.cp) {
 		/* Channel report word pending */
 		mcck->channel_report = 1;
@@ -377,6 +394,19 @@ void notrace s390_do_machine_check(struct pt_regs *regs)
 		mcck->warning = 1;
 		set_cpu_flag(CIF_MCCK_PENDING);
 	}
+
+	/*
+	 * If there are only Channel Report Pending and External Damage
+	 * machine checks, they will not be reinjected into the guest
+	 * because they refer to host conditions only.
+	 */
+	mcck_dam_code = (mci.val & MCIC_SUBCLASS_MASK);
+	if (test_cpu_flag(CIF_MCCK_GUEST) &&
+	(mcck_dam_code & MCCK_CODE_NO_GUEST) != mcck_dam_code) {
+		/* Set exit reason code for host's later handling */
+		*((long *)(regs->gprs[15] + __SF_SIE_REASON)) = -EINTR;
+	}
+	clear_cpu_flag(CIF_MCCK_GUEST);
 	nmi_exit();
 }
 
-- 
cgit 


From da72ca4d4090a8ab0e6b0a23682ef42d39d7ae00 Mon Sep 17 00:00:00 2001
From: QingFeng Hao <haoqf@linux.vnet.ibm.com>
Date: Wed, 7 Jun 2017 11:41:19 +0200
Subject: KVM: s390: Backup the guest's machine check info

When a machine check happens in the guest, related mcck info (mcic,
external damage code, ...) is stored in the vcpu's lowcore on the host.
Then the machine check handler's low-level part is executed, followed
by the high-level part.

If the high-level part's execution is interrupted by a new machine check
happening on the same vcpu on the host, the mcck info in the lowcore is
overwritten with the new machine check's data.

If the high-level part's execution is scheduled to a different cpu,
the mcck info in the lowcore is uncertain.

Therefore, for both cases, the further reinjection to the guest will use
the wrong data.
Let's backup the mcck info in the lowcore to the sie page
for further reinjection, so that the right data will be used.

Add new member into struct sie_page to store related machine check's
info of mcic, failing storage address and external damage code.

Signed-off-by: QingFeng Hao <haoqf@linux.vnet.ibm.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
 arch/s390/include/asm/kvm_host.h | 17 ++++++++++++++++-
 arch/s390/kernel/nmi.c           | 34 ++++++++++++++++++++++++++++++++++
 arch/s390/kvm/kvm-s390.c         |  1 +
 3 files changed, 51 insertions(+), 1 deletion(-)

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h
index 426614a882a9..c6e1d5fa1ad1 100644
--- a/arch/s390/include/asm/kvm_host.h
+++ b/arch/s390/include/asm/kvm_host.h
@@ -107,6 +107,20 @@ struct esca_block {
 	struct esca_entry cpu[KVM_S390_ESCA_CPU_SLOTS];
 } __packed;
 
+/*
+ * This struct is used to store some machine check info from lowcore
+ * for machine checks that happen while the guest is running.
+ * This info in host's lowcore might be overwritten by a second machine
+ * check from host when host is in the machine check's high-level handling.
+ * The size is 24 bytes.
+ */
+struct mcck_volatile_info {
+	__u64 mcic;
+	__u64 failing_storage_address;
+	__u32 ext_damage_code;
+	__u32 reserved;
+};
+
 #define CPUSTAT_STOPPED    0x80000000
 #define CPUSTAT_WAIT       0x10000000
 #define CPUSTAT_ECALL_PEND 0x08000000
@@ -264,7 +278,8 @@ struct kvm_s390_itdb {
 
 struct sie_page {
 	struct kvm_s390_sie_block sie_block;
-	__u8 reserved200[1024];		/* 0x0200 */
+	struct mcck_volatile_info mcck_info;	/* 0x0200 */
+	__u8 reserved218[1000];		/* 0x0218 */
 	struct kvm_s390_itdb itdb;	/* 0x0600 */
 	__u8 reserved700[2304];		/* 0x0700 */
 } __packed;
diff --git a/arch/s390/kernel/nmi.c b/arch/s390/kernel/nmi.c
index 958cc3352faa..31d03a84126c 100644
--- a/arch/s390/kernel/nmi.c
+++ b/arch/s390/kernel/nmi.c
@@ -26,6 +26,7 @@
 #include <asm/switch_to.h>
 #include <asm/ctl_reg.h>
 #include <asm/asm-offsets.h>
+#include <linux/kvm_host.h>
 
 struct mcck_struct {
 	unsigned int kill_task : 1;
@@ -275,6 +276,31 @@ static int notrace s390_validate_registers(union mci mci, int umode)
 	return kill_task;
 }
 
+/*
+ * Backup the guest's machine check info to its description block
+ */
+static void notrace s390_backup_mcck_info(struct pt_regs *regs)
+{
+	struct mcck_volatile_info *mcck_backup;
+	struct sie_page *sie_page;
+
+	/* r14 contains the sie block, which was set in sie64a */
+	struct kvm_s390_sie_block *sie_block =
+			(struct kvm_s390_sie_block *) regs->gprs[14];
+
+	if (sie_block == NULL)
+		/* Something's seriously wrong, stop system. */
+		s390_handle_damage();
+
+	sie_page = container_of(sie_block, struct sie_page, sie_block);
+	mcck_backup = &sie_page->mcck_info;
+	mcck_backup->mcic = S390_lowcore.mcck_interruption_code &
+				~(MCCK_CODE_CP | MCCK_CODE_EXT_DAMAGE);
+	mcck_backup->ext_damage_code = S390_lowcore.external_damage_code;
+	mcck_backup->failing_storage_address
+			= S390_lowcore.failing_storage_address;
+}
+
 #define MAX_IPD_COUNT	29
 #define MAX_IPD_TIME	(5 * 60 * USEC_PER_SEC) /* 5 minutes */
 
@@ -355,6 +381,14 @@ void notrace s390_do_machine_check(struct pt_regs *regs)
 		mcck->mcck_code = mci.val;
 		set_cpu_flag(CIF_MCCK_PENDING);
 	}
+
+	/*
+	 * Backup the machine check's info if it happens when the guest
+	 * is running.
+	 */
+	if (test_cpu_flag(CIF_MCCK_GUEST))
+		s390_backup_mcck_info(regs);
+
 	if (mci.cd) {
 		/* Timing facility damage */
 		s390_handle_damage();
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 689ac48361c6..0457e03199c5 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2069,6 +2069,7 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm,
 	if (!vcpu)
 		goto out;
 
+	BUILD_BUG_ON(sizeof(struct sie_page) != 4096);
 	sie_page = (struct sie_page *) get_zeroed_page(GFP_KERNEL);
 	if (!sie_page)
 		goto out_free_cpu;
-- 
cgit 


From 795818e8bf17dbc791764ba9fb723278ee934676 Mon Sep 17 00:00:00 2001
From: Sebastian Ott <sebott@linux.vnet.ibm.com>
Date: Sat, 10 Jun 2017 13:09:07 +0200
Subject: s390/pci: don't cleanup in arch_setup_msi_irqs

After failures in arch_setup_msi_irqs common code calls
arch_teardown_msi_irqs. Thus, remove cleanup code from
arch_setup_msi_irqs.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Reviewed-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/include/asm/pci.h |  2 +-
 arch/s390/pci/pci.c         | 45 +++++++++++++++++----------------------------
 2 files changed, 18 insertions(+), 29 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/pci.h b/arch/s390/include/asm/pci.h
index 4e3186649578..328142c8fe92 100644
--- a/arch/s390/include/asm/pci.h
+++ b/arch/s390/include/asm/pci.h
@@ -109,7 +109,7 @@ struct zpci_dev {
 	u64		msi_addr;	/* MSI address */
 	unsigned int	max_msi;	/* maximum number of MSI's */
 	struct airq_iv *aibv;		/* adapter interrupt bit vector */
-	unsigned int	aisb;		/* number of the summary bit */
+	unsigned long	aisb;		/* number of the summary bit */
 
 	/* DMA stuff */
 	unsigned long	*dma_table;
diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c
index 8051df109db3..3dd9686c576a 100644
--- a/arch/s390/pci/pci.c
+++ b/arch/s390/pci/pci.c
@@ -372,22 +372,21 @@ int arch_setup_msi_irqs(struct pci_dev *pdev, int nvec, int type)
 	struct msi_msg msg;
 	int rc, irq;
 
+	zdev->aisb = -1UL;
 	if (type == PCI_CAP_ID_MSI && nvec > 1)
 		return 1;
 	msi_vecs = min_t(unsigned int, nvec, zdev->max_msi);
 
 	/* Allocate adapter summary indicator bit */
-	rc = -EIO;
 	aisb = airq_iv_alloc_bit(zpci_aisb_iv);
 	if (aisb == -1UL)
-		goto out;
+		return -EIO;
 	zdev->aisb = aisb;
 
 	/* Create adapter interrupt vector */
-	rc = -ENOMEM;
 	zdev->aibv = airq_iv_create(msi_vecs, AIRQ_IV_DATA | AIRQ_IV_BITLOCK);
 	if (!zdev->aibv)
-		goto out_si;
+		return -ENOMEM;
 
 	/* Wire up shortcut pointer */
 	zpci_aibv[aisb] = zdev->aibv;
@@ -398,10 +397,10 @@ int arch_setup_msi_irqs(struct pci_dev *pdev, int nvec, int type)
 		rc = -EIO;
 		irq = irq_alloc_desc(0);	/* Alloc irq on node 0 */
 		if (irq < 0)
-			goto out_msi;
+			return -ENOMEM;
 		rc = irq_set_msi_desc(irq, msi);
 		if (rc)
-			goto out_msi;
+			return rc;
 		irq_set_chip_and_handler(irq, &zpci_irq_chip,
 					 handle_simple_irq);
 		msg.data = hwirq;
@@ -415,27 +414,9 @@ int arch_setup_msi_irqs(struct pci_dev *pdev, int nvec, int type)
 	/* Enable adapter interrupts */
 	rc = zpci_set_airq(zdev);
 	if (rc)
-		goto out_msi;
+		return rc;
 
 	return (msi_vecs == nvec) ? 0 : msi_vecs;
-
-out_msi:
-	for_each_pci_msi_entry(msi, pdev) {
-		if (hwirq-- == 0)
-			break;
-		irq_set_msi_desc(msi->irq, NULL);
-		irq_free_desc(msi->irq);
-		msi->msg.address_lo = 0;
-		msi->msg.address_hi = 0;
-		msi->msg.data = 0;
-		msi->irq = 0;
-	}
-	zpci_aibv[aisb] = NULL;
-	airq_iv_release(zdev->aibv);
-out_si:
-	airq_iv_free_bit(zpci_aisb_iv, aisb);
-out:
-	return rc;
 }
 
 void arch_teardown_msi_irqs(struct pci_dev *pdev)
@@ -451,6 +432,8 @@ void arch_teardown_msi_irqs(struct pci_dev *pdev)
 
 	/* Release MSI interrupts */
 	for_each_pci_msi_entry(msi, pdev) {
+		if (!msi->irq)
+			continue;
 		if (msi->msi_attrib.is_msix)
 			__pci_msix_desc_mask_irq(msi, 1);
 		else
@@ -463,9 +446,15 @@ void arch_teardown_msi_irqs(struct pci_dev *pdev)
 		msi->irq = 0;
 	}
 
-	zpci_aibv[zdev->aisb] = NULL;
-	airq_iv_release(zdev->aibv);
-	airq_iv_free_bit(zpci_aisb_iv, zdev->aisb);
+	if (zdev->aisb != -1UL) {
+		zpci_aibv[zdev->aisb] = NULL;
+		airq_iv_free_bit(zpci_aisb_iv, zdev->aisb);
+		zdev->aisb = -1UL;
+	}
+	if (zdev->aibv) {
+		airq_iv_release(zdev->aibv);
+		zdev->aibv = NULL;
+	}
 }
 
 static void zpci_map_resources(struct pci_dev *pdev)
-- 
cgit 


From 4dfbd3efe3f0cf9ff1325b87491e1b1fe07afaf1 Mon Sep 17 00:00:00 2001
From: Sebastian Ott <sebott@linux.vnet.ibm.com>
Date: Sat, 10 Jun 2017 13:54:44 +0200
Subject: s390/pci: improve error handling during interrupt deregistration

When we ask a function to stop creating interrupts this may fail
due to the function being already gone (e.g. after hot-unplug).

Consequently we don't free associated resources like summary bits
and bit vectors used for irq processing. This could lead to
situations where we ran out of these resources and fail to setup
new interrupts.

The fix is to just ignore the errors in cases where we can be
sure no new interrupts are generated.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Reviewed-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/include/asm/pci_insn.h |  2 +-
 arch/s390/pci/pci.c              | 29 +++++++++++++++++++----------
 arch/s390/pci/pci_insn.c         | 10 +++++-----
 3 files changed, 25 insertions(+), 16 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/pci_insn.h b/arch/s390/include/asm/pci_insn.h
index 649eb62c52b3..34abcf275799 100644
--- a/arch/s390/include/asm/pci_insn.h
+++ b/arch/s390/include/asm/pci_insn.h
@@ -76,7 +76,7 @@ struct zpci_fib {
 	u32 gd;
 } __packed __aligned(8);
 
-int zpci_mod_fc(u64 req, struct zpci_fib *fib);
+u8 zpci_mod_fc(u64 req, struct zpci_fib *fib, u8 *status);
 int zpci_refresh_trans(u64 fn, u64 addr, u64 range);
 int zpci_load(u64 *data, u64 req, u64 offset);
 int zpci_store(u64 data, u64 req, u64 offset);
diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c
index 3dd9686c576a..82f3e788b0be 100644
--- a/arch/s390/pci/pci.c
+++ b/arch/s390/pci/pci.c
@@ -108,6 +108,7 @@ static int zpci_set_airq(struct zpci_dev *zdev)
 {
 	u64 req = ZPCI_CREATE_REQ(zdev->fh, 0, ZPCI_MOD_FC_REG_INT);
 	struct zpci_fib fib = {0};
+	u8 status;
 
 	fib.isc = PCI_ISC;
 	fib.sum = 1;		/* enable summary notifications */
@@ -117,7 +118,22 @@ static int zpci_set_airq(struct zpci_dev *zdev)
 	fib.aisb = (unsigned long) zpci_aisb_iv->vector + (zdev->aisb/64)*8;
 	fib.aisbo = zdev->aisb & 63;
 
-	return zpci_mod_fc(req, &fib);
+	return zpci_mod_fc(req, &fib, &status) ? -EIO : 0;
+}
+
+/* Modify PCI: Unregister adapter interruptions */
+static int zpci_clear_airq(struct zpci_dev *zdev)
+{
+	u64 req = ZPCI_CREATE_REQ(zdev->fh, 0, ZPCI_MOD_FC_DEREG_INT);
+	struct zpci_fib fib = {0};
+	u8 cc, status;
+
+	cc = zpci_mod_fc(req, &fib, &status);
+	if (cc == 3 || (cc == 1 && status == 24))
+		/* Function already gone or IRQs already deregistered. */
+		cc = 0;
+
+	return cc ? -EIO : 0;
 }
 
 struct mod_pci_args {
@@ -131,13 +147,14 @@ static int mod_pci(struct zpci_dev *zdev, int fn, u8 dmaas, struct mod_pci_args
 {
 	u64 req = ZPCI_CREATE_REQ(zdev->fh, dmaas, fn);
 	struct zpci_fib fib = {0};
+	u8 status;
 
 	fib.pba = args->base;
 	fib.pal = args->limit;
 	fib.iota = args->iota;
 	fib.fmb_addr = args->fmb_addr;
 
-	return zpci_mod_fc(req, &fib);
+	return zpci_mod_fc(req, &fib, &status) ? -EIO : 0;
 }
 
 /* Modify PCI: Register I/O address translation parameters */
@@ -159,14 +176,6 @@ int zpci_unregister_ioat(struct zpci_dev *zdev, u8 dmaas)
 	return mod_pci(zdev, ZPCI_MOD_FC_DEREG_IOAT, dmaas, &args);
 }
 
-/* Modify PCI: Unregister adapter interruptions */
-static int zpci_clear_airq(struct zpci_dev *zdev)
-{
-	struct mod_pci_args args = { 0, 0, 0, 0 };
-
-	return mod_pci(zdev, ZPCI_MOD_FC_DEREG_INT, 0, &args);
-}
-
 /* Modify PCI: Set PCI function measurement parameters */
 int zpci_fmb_enable_device(struct zpci_dev *zdev)
 {
diff --git a/arch/s390/pci/pci_insn.c b/arch/s390/pci/pci_insn.c
index fa8d7d4b9751..ea34086c8674 100644
--- a/arch/s390/pci/pci_insn.c
+++ b/arch/s390/pci/pci_insn.c
@@ -40,20 +40,20 @@ static inline u8 __mpcifc(u64 req, struct zpci_fib *fib, u8 *status)
 	return cc;
 }
 
-int zpci_mod_fc(u64 req, struct zpci_fib *fib)
+u8 zpci_mod_fc(u64 req, struct zpci_fib *fib, u8 *status)
 {
-	u8 cc, status;
+	u8 cc;
 
 	do {
-		cc = __mpcifc(req, fib, &status);
+		cc = __mpcifc(req, fib, status);
 		if (cc == 2)
 			msleep(ZPCI_INSN_BUSY_DELAY);
 	} while (cc == 2);
 
 	if (cc)
-		zpci_err_insn(cc, status, req, 0);
+		zpci_err_insn(cc, *status, req, 0);
 
-	return (cc) ? -EIO : 0;
+	return cc;
 }
 
 /* Refresh PCI Translations */
-- 
cgit 


From 725708349172f00b86e2bac5c03de360b79aaf65 Mon Sep 17 00:00:00 2001
From: Sebastian Ott <sebott@linux.vnet.ibm.com>
Date: Sat, 10 Jun 2017 14:10:00 +0200
Subject: s390/pci: improve unreg_ioat error handling

DMA tables are freed in zpci_dma_exit_device regardless of the return
code of zpci_unregister_ioat. This could lead to a use after free. On
the other hand during function hot-unplug, zpci_unregister_ioat will
always fail since the function is already gone.

So let zpci_unregister_ioat report success when the function is gone
but don't cleanup the dma table when a function could still have it
in access.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Reviewed-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/pci/pci.c     | 19 ++++++++++++++-----
 arch/s390/pci/pci_dma.c |  4 +++-
 2 files changed, 17 insertions(+), 6 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c
index 82f3e788b0be..4f541f54470f 100644
--- a/arch/s390/pci/pci.c
+++ b/arch/s390/pci/pci.c
@@ -161,19 +161,28 @@ static int mod_pci(struct zpci_dev *zdev, int fn, u8 dmaas, struct mod_pci_args
 int zpci_register_ioat(struct zpci_dev *zdev, u8 dmaas,
 		       u64 base, u64 limit, u64 iota)
 {
-	struct mod_pci_args args = { base, limit, iota, 0 };
+	u64 req = ZPCI_CREATE_REQ(zdev->fh, dmaas, ZPCI_MOD_FC_REG_IOAT);
+	struct zpci_fib fib = {0};
+	u8 status;
 
 	WARN_ON_ONCE(iota & 0x3fff);
-	args.iota |= ZPCI_IOTA_RTTO_FLAG;
-	return mod_pci(zdev, ZPCI_MOD_FC_REG_IOAT, dmaas, &args);
+	fib.pba = base;
+	fib.pal = limit;
+	fib.iota = iota | ZPCI_IOTA_RTTO_FLAG;
+	return zpci_mod_fc(req, &fib, &status) ? -EIO : 0;
 }
 
 /* Modify PCI: Unregister I/O address translation parameters */
 int zpci_unregister_ioat(struct zpci_dev *zdev, u8 dmaas)
 {
-	struct mod_pci_args args = { 0, 0, 0, 0 };
+	u64 req = ZPCI_CREATE_REQ(zdev->fh, dmaas, ZPCI_MOD_FC_DEREG_IOAT);
+	struct zpci_fib fib = {0};
+	u8 cc, status;
 
-	return mod_pci(zdev, ZPCI_MOD_FC_DEREG_IOAT, dmaas, &args);
+	cc = zpci_mod_fc(req, &fib, &status);
+	if (cc == 3) /* Function already gone. */
+		cc = 0;
+	return cc ? -EIO : 0;
 }
 
 /* Modify PCI: Set PCI function measurement parameters */
diff --git a/arch/s390/pci/pci_dma.c b/arch/s390/pci/pci_dma.c
index 9081a57fa340..8eb1cc341dab 100644
--- a/arch/s390/pci/pci_dma.c
+++ b/arch/s390/pci/pci_dma.c
@@ -601,7 +601,9 @@ void zpci_dma_exit_device(struct zpci_dev *zdev)
 	 */
 	WARN_ON(zdev->s390_domain);
 
-	zpci_unregister_ioat(zdev, 0);
+	if (zpci_unregister_ioat(zdev, 0))
+		return;
+
 	dma_cleanup_tables(zdev->dma_table);
 	zdev->dma_table = NULL;
 	vfree(zdev->iommu_bitmap);
-- 
cgit 


From 4e5bd7803bd1e822513d1007bad1527a2ec589b6 Mon Sep 17 00:00:00 2001
From: Sebastian Ott <sebott@linux.vnet.ibm.com>
Date: Sat, 10 Jun 2017 14:12:13 +0200
Subject: s390/pci: improve error handling during fmb (de)registration

Cleanup in zpci_fmb_enable_device when fmb registration fails. Also
don't free the fmb when deregistration fails in zpci_fmb_disable_device
but handle error situations when a function was hot-unplugged.

Also remove the mod_pci helper since it is no longer used.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Reviewed-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/pci/pci.c | 51 +++++++++++++++++++++------------------------------
 1 file changed, 21 insertions(+), 30 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c
index 4f541f54470f..6a44a68efb81 100644
--- a/arch/s390/pci/pci.c
+++ b/arch/s390/pci/pci.c
@@ -136,27 +136,6 @@ static int zpci_clear_airq(struct zpci_dev *zdev)
 	return cc ? -EIO : 0;
 }
 
-struct mod_pci_args {
-	u64 base;
-	u64 limit;
-	u64 iota;
-	u64 fmb_addr;
-};
-
-static int mod_pci(struct zpci_dev *zdev, int fn, u8 dmaas, struct mod_pci_args *args)
-{
-	u64 req = ZPCI_CREATE_REQ(zdev->fh, dmaas, fn);
-	struct zpci_fib fib = {0};
-	u8 status;
-
-	fib.pba = args->base;
-	fib.pal = args->limit;
-	fib.iota = args->iota;
-	fib.fmb_addr = args->fmb_addr;
-
-	return zpci_mod_fc(req, &fib, &status) ? -EIO : 0;
-}
-
 /* Modify PCI: Register I/O address translation parameters */
 int zpci_register_ioat(struct zpci_dev *zdev, u8 dmaas,
 		       u64 base, u64 limit, u64 iota)
@@ -188,7 +167,9 @@ int zpci_unregister_ioat(struct zpci_dev *zdev, u8 dmaas)
 /* Modify PCI: Set PCI function measurement parameters */
 int zpci_fmb_enable_device(struct zpci_dev *zdev)
 {
-	struct mod_pci_args args = { 0, 0, 0, 0 };
+	u64 req = ZPCI_CREATE_REQ(zdev->fh, 0, ZPCI_MOD_FC_SET_MEASURE);
+	struct zpci_fib fib = {0};
+	u8 cc, status;
 
 	if (zdev->fmb || sizeof(*zdev->fmb) < zdev->fmb_length)
 		return -EINVAL;
@@ -203,25 +184,35 @@ int zpci_fmb_enable_device(struct zpci_dev *zdev)
 	atomic64_set(&zdev->mapped_pages, 0);
 	atomic64_set(&zdev->unmapped_pages, 0);
 
-	args.fmb_addr = virt_to_phys(zdev->fmb);
-	return mod_pci(zdev, ZPCI_MOD_FC_SET_MEASURE, 0, &args);
+	fib.fmb_addr = virt_to_phys(zdev->fmb);
+	cc = zpci_mod_fc(req, &fib, &status);
+	if (cc) {
+		kmem_cache_free(zdev_fmb_cache, zdev->fmb);
+		zdev->fmb = NULL;
+	}
+	return cc ? -EIO : 0;
 }
 
 /* Modify PCI: Disable PCI function measurement */
 int zpci_fmb_disable_device(struct zpci_dev *zdev)
 {
-	struct mod_pci_args args = { 0, 0, 0, 0 };
-	int rc;
+	u64 req = ZPCI_CREATE_REQ(zdev->fh, 0, ZPCI_MOD_FC_SET_MEASURE);
+	struct zpci_fib fib = {0};
+	u8 cc, status;
 
 	if (!zdev->fmb)
 		return -EINVAL;
 
 	/* Function measurement is disabled if fmb address is zero */
-	rc = mod_pci(zdev, ZPCI_MOD_FC_SET_MEASURE, 0, &args);
+	cc = zpci_mod_fc(req, &fib, &status);
+	if (cc == 3) /* Function already gone. */
+		cc = 0;
 
-	kmem_cache_free(zdev_fmb_cache, zdev->fmb);
-	zdev->fmb = NULL;
-	return rc;
+	if (!cc) {
+		kmem_cache_free(zdev_fmb_cache, zdev->fmb);
+		zdev->fmb = NULL;
+	}
+	return cc ? -EIO : 0;
 }
 
 static int zpci_cfg_load(struct zpci_dev *zdev, int offset, u32 *val, u8 len)
-- 
cgit 


From 783684f1f60faec09f3ac74c0b12e89bdb182429 Mon Sep 17 00:00:00 2001
From: Sebastian Ott <sebott@linux.vnet.ibm.com>
Date: Wed, 26 Apr 2017 18:59:52 +0200
Subject: s390/pci: introduce clp_get_state

Code handling pci hotplug needs to determine the configuration
state of a pci function. Implement clp_get_state as a wrapper
for list pci functions.

Also change enum zpci_state to match the configuration state
values.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Reviewed-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/include/asm/pci.h | 10 ++++-----
 arch/s390/pci/pci_clp.c     | 51 +++++++++++++++++++++++++++++++++++++--------
 2 files changed, 47 insertions(+), 14 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/pci.h b/arch/s390/include/asm/pci.h
index 328142c8fe92..01c58d41bee8 100644
--- a/arch/s390/include/asm/pci.h
+++ b/arch/s390/include/asm/pci.h
@@ -70,11 +70,10 @@ struct zpci_fmb {
 } __packed __aligned(128);
 
 enum zpci_state {
-	ZPCI_FN_STATE_RESERVED,
-	ZPCI_FN_STATE_STANDBY,
-	ZPCI_FN_STATE_CONFIGURED,
-	ZPCI_FN_STATE_ONLINE,
-	NR_ZPCI_FN_STATES,
+	ZPCI_FN_STATE_STANDBY = 0,
+	ZPCI_FN_STATE_CONFIGURED = 1,
+	ZPCI_FN_STATE_RESERVED = 2,
+	ZPCI_FN_STATE_ONLINE = 3,
 };
 
 struct zpci_bar_struct {
@@ -172,6 +171,7 @@ int clp_rescan_pci_devices_simple(void);
 int clp_add_pci_device(u32, u32, int);
 int clp_enable_fh(struct zpci_dev *, u8);
 int clp_disable_fh(struct zpci_dev *);
+int clp_get_state(u32 fid, enum zpci_state *state);
 
 #ifdef CONFIG_PCI
 /* Error handling and recovery */
diff --git a/arch/s390/pci/pci_clp.c b/arch/s390/pci/pci_clp.c
index 1c3332ac1957..267cdd69e6da 100644
--- a/arch/s390/pci/pci_clp.c
+++ b/arch/s390/pci/pci_clp.c
@@ -295,8 +295,8 @@ int clp_disable_fh(struct zpci_dev *zdev)
 	return rc;
 }
 
-static int clp_list_pci(struct clp_req_rsp_list_pci *rrb,
-			void (*cb)(struct clp_fh_list_entry *entry))
+static int clp_list_pci(struct clp_req_rsp_list_pci *rrb, void *data,
+			void (*cb)(struct clp_fh_list_entry *, void *))
 {
 	u64 resume_token = 0;
 	int entries, i, rc;
@@ -327,13 +327,13 @@ static int clp_list_pci(struct clp_req_rsp_list_pci *rrb,
 
 		resume_token = rrb->response.resume_token;
 		for (i = 0; i < entries; i++)
-			cb(&rrb->response.fh_list[i]);
+			cb(&rrb->response.fh_list[i], data);
 	} while (resume_token);
 out:
 	return rc;
 }
 
-static void __clp_add(struct clp_fh_list_entry *entry)
+static void __clp_add(struct clp_fh_list_entry *entry, void *data)
 {
 	if (!entry->vendor_id)
 		return;
@@ -341,7 +341,7 @@ static void __clp_add(struct clp_fh_list_entry *entry)
 	clp_add_pci_device(entry->fid, entry->fh, entry->config_state);
 }
 
-static void __clp_rescan(struct clp_fh_list_entry *entry)
+static void __clp_rescan(struct clp_fh_list_entry *entry, void *data)
 {
 	struct zpci_dev *zdev;
 
@@ -364,7 +364,7 @@ static void __clp_rescan(struct clp_fh_list_entry *entry)
 	}
 }
 
-static void __clp_update(struct clp_fh_list_entry *entry)
+static void __clp_update(struct clp_fh_list_entry *entry, void *data)
 {
 	struct zpci_dev *zdev;
 
@@ -387,7 +387,7 @@ int clp_scan_pci_devices(void)
 	if (!rrb)
 		return -ENOMEM;
 
-	rc = clp_list_pci(rrb, __clp_add);
+	rc = clp_list_pci(rrb, NULL, __clp_add);
 
 	clp_free_block(rrb);
 	return rc;
@@ -402,7 +402,7 @@ int clp_rescan_pci_devices(void)
 	if (!rrb)
 		return -ENOMEM;
 
-	rc = clp_list_pci(rrb, __clp_rescan);
+	rc = clp_list_pci(rrb, NULL, __clp_rescan);
 
 	clp_free_block(rrb);
 	return rc;
@@ -417,7 +417,40 @@ int clp_rescan_pci_devices_simple(void)
 	if (!rrb)
 		return -ENOMEM;
 
-	rc = clp_list_pci(rrb, __clp_update);
+	rc = clp_list_pci(rrb, NULL, __clp_update);
+
+	clp_free_block(rrb);
+	return rc;
+}
+
+struct clp_state_data {
+	u32 fid;
+	enum zpci_state state;
+};
+
+static void __clp_get_state(struct clp_fh_list_entry *entry, void *data)
+{
+	struct clp_state_data *sd = data;
+
+	if (entry->fid != sd->fid)
+		return;
+
+	sd->state = entry->config_state;
+}
+
+int clp_get_state(u32 fid, enum zpci_state *state)
+{
+	struct clp_req_rsp_list_pci *rrb;
+	struct clp_state_data sd = {fid, ZPCI_FN_STATE_RESERVED};
+	int rc;
+
+	rrb = clp_alloc_block(GFP_KERNEL);
+	if (!rrb)
+		return -ENOMEM;
+
+	rc = clp_list_pci(rrb, &sd, __clp_get_state);
+	if (!rc)
+		*state = sd.state;
 
 	clp_free_block(rrb);
 	return rc;
-- 
cgit 


From 623bd44d3f277b7bbe16e0e091bd361e75964b5d Mon Sep 17 00:00:00 2001
From: Sebastian Ott <sebott@linux.vnet.ibm.com>
Date: Tue, 9 May 2017 12:27:30 +0200
Subject: s390/pci: improve pci hotplug

PCI hotplug events basically notify about the new state of a
function. Unfortunately some hypervisors implement hotplug
events in a way where it is not clear what the new state of
the function should be.

Use clp_get_state to find the current state of the function
and handle accordingly.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Reviewed-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/include/asm/pci.h |  1 +
 arch/s390/pci/pci.c         |  9 +++++++++
 arch/s390/pci/pci_event.c   | 14 +++++++++++---
 3 files changed, 21 insertions(+), 3 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/pci.h b/arch/s390/include/asm/pci.h
index 01c58d41bee8..280458c82104 100644
--- a/arch/s390/include/asm/pci.h
+++ b/arch/s390/include/asm/pci.h
@@ -158,6 +158,7 @@ extern const struct attribute_group *zpci_attr_groups[];
 ----------------------------------------------------------------------------- */
 /* Base stuff */
 int zpci_create_device(struct zpci_dev *);
+void zpci_remove_device(struct zpci_dev *zdev);
 int zpci_enable_device(struct zpci_dev *);
 int zpci_disable_device(struct zpci_dev *);
 void zpci_stop_device(struct zpci_dev *);
diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c
index 6a44a68efb81..f4928bc57773 100644
--- a/arch/s390/pci/pci.c
+++ b/arch/s390/pci/pci.c
@@ -855,6 +855,15 @@ void zpci_stop_device(struct zpci_dev *zdev)
 }
 EXPORT_SYMBOL_GPL(zpci_stop_device);
 
+void zpci_remove_device(struct zpci_dev *zdev)
+{
+	if (!zdev->bus)
+		return;
+
+	pci_stop_root_bus(zdev->bus);
+	pci_remove_root_bus(zdev->bus);
+}
+
 int zpci_report_error(struct pci_dev *pdev,
 		      struct zpci_report_error_header *report)
 {
diff --git a/arch/s390/pci/pci_event.c b/arch/s390/pci/pci_event.c
index c2b27ad8e94d..0bbc04af4418 100644
--- a/arch/s390/pci/pci_event.c
+++ b/arch/s390/pci/pci_event.c
@@ -74,6 +74,7 @@ static void __zpci_event_availability(struct zpci_ccdf_avail *ccdf)
 {
 	struct zpci_dev *zdev = get_zdev_by_fid(ccdf->fid);
 	struct pci_dev *pdev = NULL;
+	enum zpci_state state;
 	int ret;
 
 	if (zdev)
@@ -108,6 +109,8 @@ static void __zpci_event_availability(struct zpci_ccdf_avail *ccdf)
 			clp_add_pci_device(ccdf->fid, ccdf->fh, 0);
 		break;
 	case 0x0303: /* Deconfiguration requested */
+		if (!zdev)
+			break;
 		if (pdev)
 			pci_stop_and_remove_bus_device_locked(pdev);
 
@@ -121,7 +124,9 @@ static void __zpci_event_availability(struct zpci_ccdf_avail *ccdf)
 			zdev->state = ZPCI_FN_STATE_STANDBY;
 
 		break;
-	case 0x0304: /* Configured -> Standby */
+	case 0x0304: /* Configured -> Standby|Reserved */
+		if (!zdev)
+			break;
 		if (pdev) {
 			/* Give the driver a hint that the function is
 			 * already unusable. */
@@ -132,6 +137,10 @@ static void __zpci_event_availability(struct zpci_ccdf_avail *ccdf)
 		zdev->fh = ccdf->fh;
 		zpci_disable_device(zdev);
 		zdev->state = ZPCI_FN_STATE_STANDBY;
+		if (!clp_get_state(ccdf->fid, &state) &&
+		    state == ZPCI_FN_STATE_RESERVED) {
+			zpci_remove_device(zdev);
+		}
 		break;
 	case 0x0306: /* 0x308 or 0x302 for multiple devices */
 		clp_rescan_pci_devices();
@@ -139,8 +148,7 @@ static void __zpci_event_availability(struct zpci_ccdf_avail *ccdf)
 	case 0x0308: /* Standby -> Reserved */
 		if (!zdev)
 			break;
-		pci_stop_root_bus(zdev->bus);
-		pci_remove_root_bus(zdev->bus);
+		zpci_remove_device(zdev);
 		break;
 	default:
 		break;
-- 
cgit 


From 01553d9a2ba7c658bf9e9e5c65466508c1bd6db5 Mon Sep 17 00:00:00 2001
From: Sebastian Ott <sebott@linux.vnet.ibm.com>
Date: Tue, 20 Jun 2017 15:56:05 +0200
Subject: s390/pci: fix handling of PEC 306

In contrast to other hotplug events PEC 0x306 isn't about a single
but multiple devices. Also there's no information on what happened
to these devices. We correctly handled hotplug that way but failed
to handle hot-unplug. This patch addresses that and implements
hot-unplug of multiple devices via PEC 306.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Reviewed-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/include/asm/pci.h |  2 +-
 arch/s390/pci/pci.c         | 29 +++++++++++++++++++----------
 arch/s390/pci/pci_clp.c     | 25 ++++---------------------
 3 files changed, 24 insertions(+), 32 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/include/asm/pci.h b/arch/s390/include/asm/pci.h
index 280458c82104..f36b4b726057 100644
--- a/arch/s390/include/asm/pci.h
+++ b/arch/s390/include/asm/pci.h
@@ -161,9 +161,9 @@ int zpci_create_device(struct zpci_dev *);
 void zpci_remove_device(struct zpci_dev *zdev);
 int zpci_enable_device(struct zpci_dev *);
 int zpci_disable_device(struct zpci_dev *);
-void zpci_stop_device(struct zpci_dev *);
 int zpci_register_ioat(struct zpci_dev *, u8, u64, u64, u64);
 int zpci_unregister_ioat(struct zpci_dev *, u8);
+void zpci_remove_reserved_devices(void);
 
 /* CLP */
 int clp_scan_pci_devices(void);
diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c
index f4928bc57773..114b390d80f9 100644
--- a/arch/s390/pci/pci.c
+++ b/arch/s390/pci/pci.c
@@ -86,6 +86,25 @@ struct zpci_dev *get_zdev_by_fid(u32 fid)
 	return zdev;
 }
 
+void zpci_remove_reserved_devices(void)
+{
+	struct zpci_dev *tmp, *zdev;
+	enum zpci_state state;
+	LIST_HEAD(remove);
+
+	spin_lock(&zpci_list_lock);
+	list_for_each_entry_safe(zdev, tmp, &zpci_list, entry) {
+		if (zdev->state == ZPCI_FN_STATE_STANDBY &&
+		    !clp_get_state(zdev->fid, &state) &&
+		    state == ZPCI_FN_STATE_RESERVED)
+			list_move_tail(&zdev->entry, &remove);
+	}
+	spin_unlock(&zpci_list_lock);
+
+	list_for_each_entry_safe(zdev, tmp, &remove, entry)
+		zpci_remove_device(zdev);
+}
+
 static struct zpci_dev *get_zdev_by_bus(struct pci_bus *bus)
 {
 	return (bus && bus->sysdata) ? (struct zpci_dev *) bus->sysdata : NULL;
@@ -845,16 +864,6 @@ out:
 	return rc;
 }
 
-void zpci_stop_device(struct zpci_dev *zdev)
-{
-	zpci_dma_exit_device(zdev);
-	/*
-	 * Note: SCLP disables fh via set-pci-fn so don't
-	 * do that here.
-	 */
-}
-EXPORT_SYMBOL_GPL(zpci_stop_device);
-
 void zpci_remove_device(struct zpci_dev *zdev)
 {
 	if (!zdev->bus)
diff --git a/arch/s390/pci/pci_clp.c b/arch/s390/pci/pci_clp.c
index 267cdd69e6da..3a5cd84e5a3b 100644
--- a/arch/s390/pci/pci_clp.c
+++ b/arch/s390/pci/pci_clp.c
@@ -334,14 +334,6 @@ out:
 }
 
 static void __clp_add(struct clp_fh_list_entry *entry, void *data)
-{
-	if (!entry->vendor_id)
-		return;
-
-	clp_add_pci_device(entry->fid, entry->fh, entry->config_state);
-}
-
-static void __clp_rescan(struct clp_fh_list_entry *entry, void *data)
 {
 	struct zpci_dev *zdev;
 
@@ -349,19 +341,8 @@ static void __clp_rescan(struct clp_fh_list_entry *entry, void *data)
 		return;
 
 	zdev = get_zdev_by_fid(entry->fid);
-	if (!zdev) {
+	if (!zdev)
 		clp_add_pci_device(entry->fid, entry->fh, entry->config_state);
-		return;
-	}
-
-	if (!entry->config_state) {
-		/*
-		 * The handle is already disabled, that means no iota/irq freeing via
-		 * the firmware interfaces anymore. Need to free resources manually
-		 * (DMA memory, debug, sysfs)...
-		 */
-		zpci_stop_device(zdev);
-	}
 }
 
 static void __clp_update(struct clp_fh_list_entry *entry, void *data)
@@ -398,11 +379,13 @@ int clp_rescan_pci_devices(void)
 	struct clp_req_rsp_list_pci *rrb;
 	int rc;
 
+	zpci_remove_reserved_devices();
+
 	rrb = clp_alloc_block(GFP_KERNEL);
 	if (!rrb)
 		return -ENOMEM;
 
-	rc = clp_list_pci(rrb, NULL, __clp_rescan);
+	rc = clp_list_pci(rrb, NULL, __clp_add);
 
 	clp_free_block(rrb);
 	return rc;
-- 
cgit 


From be2c36769f6340322c5367586adad6c3ce1ed51a Mon Sep 17 00:00:00 2001
From: Sebastian Ott <sebott@linux.vnet.ibm.com>
Date: Wed, 21 Jun 2017 12:30:40 +0200
Subject: s390/pci: provide more debug information

Add some debug data to observe the lifetime of the
architecture specific device information.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Reviewed-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/pci/pci.c     | 1 +
 arch/s390/pci/pci_clp.c | 5 +++--
 2 files changed, 4 insertions(+), 2 deletions(-)

(limited to 'arch/s390')

diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c
index 114b390d80f9..d5c7444a0c05 100644
--- a/arch/s390/pci/pci.c
+++ b/arch/s390/pci/pci.c
@@ -772,6 +772,7 @@ void pcibios_remove_bus(struct pci_bus *bus)
 	list_del(&zdev->entry);
 	spin_unlock(&zpci_list_lock);
 
+	zpci_dbg(3, "rem fid:%x\n", zdev->fid);
 	kfree(zdev);
 }
 
diff --git a/arch/s390/pci/pci_clp.c b/arch/s390/pci/pci_clp.c
index 3a5cd84e5a3b..bd534b4d40e3 100644
--- a/arch/s390/pci/pci_clp.c
+++ b/arch/s390/pci/pci_clp.c
@@ -193,12 +193,12 @@ out:
 int clp_add_pci_device(u32 fid, u32 fh, int configured)
 {
 	struct zpci_dev *zdev;
-	int rc;
+	int rc = -ENOMEM;
 
 	zpci_dbg(3, "add fid:%x, fh:%x, c:%d\n", fid, fh, configured);
 	zdev = kzalloc(sizeof(*zdev), GFP_KERNEL);
 	if (!zdev)
-		return -ENOMEM;
+		goto error;
 
 	zdev->fh = fh;
 	zdev->fid = fid;
@@ -219,6 +219,7 @@ int clp_add_pci_device(u32 fid, u32 fh, int configured)
 	return 0;
 
 error:
+	zpci_dbg(0, "add fid:%x, rc:%d\n", fid, rc);
 	kfree(zdev);
 	return rc;
 }
-- 
cgit 


From 312e8462ab6a1594a0bc4bc51c8e36c9f07b447b Mon Sep 17 00:00:00 2001
From: Sebastian Ott <sebott@linux.vnet.ibm.com>
Date: Wed, 21 Jun 2017 10:20:35 +0200
Subject: s390/pci: recognize name clashes with uids

When uid checking is enabled firmware guarantees uniqueness of the uids
and we use them for device enumeration. Tests have shown that uid checking
can be toggled at runtime. This is unfortunate since it can lead to name
clashes.

Recognize these name clashes by allocating bits in zpci_domain even for
firmware provided ids.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Reviewed-by: Pierre Morel <pmorel@linux.vnet.ibm.com>
Reviewed-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/pci/pci.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

(limited to 'arch/s390')

diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c
index d5c7444a0c05..7b30af5da222 100644
--- a/arch/s390/pci/pci.c
+++ b/arch/s390/pci/pci.c
@@ -736,6 +736,16 @@ static int zpci_alloc_domain(struct zpci_dev *zdev)
 {
 	if (zpci_unique_uid) {
 		zdev->domain = (u16) zdev->uid;
+		if (zdev->domain >= ZPCI_NR_DEVICES)
+			return 0;
+
+		spin_lock(&zpci_domain_lock);
+		if (test_bit(zdev->domain, zpci_domain)) {
+			spin_unlock(&zpci_domain_lock);
+			return -EEXIST;
+		}
+		set_bit(zdev->domain, zpci_domain);
+		spin_unlock(&zpci_domain_lock);
 		return 0;
 	}
 
@@ -752,7 +762,7 @@ static int zpci_alloc_domain(struct zpci_dev *zdev)
 
 static void zpci_free_domain(struct zpci_dev *zdev)
 {
-	if (zpci_unique_uid)
+	if (zdev->domain >= ZPCI_NR_DEVICES)
 		return;
 
 	spin_lock(&zpci_domain_lock);
-- 
cgit