12 files changed, 82 insertions, 69 deletions
diff --git a/arch/sparc/include/asm/atomic_32.h b/arch/sparc/include/asm/atomic_32.h
index 58fa8a5ff0cb..7aed2be45b44 100644
--- a/arch/sparc/include/asm/atomic_32.h
+++ b/arch/sparc/include/asm/atomic_32.h
@@ -14,6 +14,7 @@
 #include <linux/types.h>
 
 #include <asm/cmpxchg.h>
+#include <asm/barrier.h>
 #include <asm-generic/atomic64.h>
 
 
@@ -52,10 +53,4 @@ void atomic_set(atomic_t *, int);
 #define atomic_dec_and_test(v) (atomic_dec_return(v) == 0)
 #define atomic_sub_and_test(i, v) (atomic_sub_return(i, v) == 0)
 
-/* Atomic operations are already serializing */
-#define smp_mb__before_atomic_dec()	barrier()
-#define smp_mb__after_atomic_dec()	barrier()
-#define smp_mb__before_atomic_inc()	barrier()
-#define smp_mb__after_atomic_inc()	barrier()
-
 #endif /* !(__ARCH_SPARC_ATOMIC__) */
diff --git a/arch/sparc/include/asm/atomic_64.h b/arch/sparc/include/asm/atomic_64.h
index 0582c02a6b0c..bb894c8bec56 100644
--- a/arch/sparc/include/asm/atomic_64.h
+++ b/arch/sparc/include/asm/atomic_64.h
@@ -9,6 +9,7 @@
 
 #include <linux/types.h>
 #include <asm/cmpxchg.h>
+#include <asm/barrier.h>
 
 #define ATOMIC_INIT(i)		{ (i) }
 #define ATOMIC64_INIT(i)	{ (i) }
@@ -108,10 +109,4 @@ static inline long atomic64_add_unless(atomic64_t *v, long a, long u)
 
 long atomic64_dec_if_positive(atomic64_t *v);
 
-/* Atomic operations are already serializing */
-#define smp_mb__before_atomic_dec()	barrier()
-#define smp_mb__after_atomic_dec()	barrier()
-#define smp_mb__before_atomic_inc()	barrier()
-#define smp_mb__after_atomic_inc()	barrier()
-
 #endif /* !(__ARCH_SPARC64_ATOMIC__) */
diff --git a/arch/sparc/include/asm/barrier_64.h b/arch/sparc/include/asm/barrier_64.h
index b5aad964558e..305dcc3dc721 100644
--- a/arch/sparc/include/asm/barrier_64.h
+++ b/arch/sparc/include/asm/barrier_64.h
@@ -68,4 +68,7 @@ do {									\
 	___p1;								\
 })
 
+#define smp_mb__before_atomic()	barrier()
+#define smp_mb__after_atomic()	barrier()
+
 #endif /* !(__SPARC64_BARRIER_H) */
diff --git a/arch/sparc/include/asm/bitops_32.h b/arch/sparc/include/asm/bitops_32.h
index 4e94825fb5a2..600ed1d9c8c8 100644
--- a/arch/sparc/include/asm/bitops_32.h
+++ b/arch/sparc/include/asm/bitops_32.h
@@ -90,9 +90,6 @@ static inline void change_bit(unsigned long nr, volatile unsigned long *addr)
 
 #include <asm-generic/bitops/non-atomic.h>
 
-#define smp_mb__before_clear_bit()	do { } while(0)
-#define smp_mb__after_clear_bit()	do { } while(0)
-
 #include <asm-generic/bitops/ffz.h>
 #include <asm-generic/bitops/__ffs.h>
 #include <asm-generic/bitops/sched.h>
diff --git a/arch/sparc/include/asm/bitops_64.h b/arch/sparc/include/asm/bitops_64.h
index 0135028e4829..2d522402a937 100644
--- a/arch/sparc/include/asm/bitops_64.h
+++ b/arch/sparc/include/asm/bitops_64.h
@@ -13,6 +13,7 @@
 
 #include <linux/compiler.h>
 #include <asm/byteorder.h>
+#include <asm/barrier.h>
 
 int test_and_set_bit(unsigned long nr, volatile unsigned long *addr);
 int test_and_clear_bit(unsigned long nr, volatile unsigned long *addr);
@@ -23,9 +24,6 @@ void change_bit(unsigned long nr, volatile unsigned long *addr);
 
 #include <asm-generic/bitops/non-atomic.h>
 
-#define smp_mb__before_clear_bit()	barrier()
-#define smp_mb__after_clear_bit()	barrier()
-
 #include <asm-generic/bitops/fls.h>
 #include <asm-generic/bitops/__fls.h>
 #include <asm-generic/bitops/fls64.h>
diff --git a/arch/sparc/include/asm/checksum_32.h b/arch/sparc/include/asm/checksum_32.h
index c80cf02079ec..426b2389a1c2 100644
--- a/arch/sparc/include/asm/checksum_32.h
+++ b/arch/sparc/include/asm/checksum_32.h
@@ -238,4 +238,16 @@ static inline __sum16 ip_compute_csum(const void *buff, int len)
 	return csum_fold(csum_partial(buff, len, 0));
 }
 
+#define HAVE_ARCH_CSUM_ADD
+static inline __wsum csum_add(__wsum csum, __wsum addend)
+{
+	__asm__ __volatile__(
+		"addcc   %0, %1, %0\n"
+		"addx    %0, %%g0, %0"
+		: "=r" (csum)
+		: "r" (addend), "0" (csum));
+
+	return csum;
+}
+
 #endif /* !(__SPARC_CHECKSUM_H) */
diff --git a/arch/sparc/include/asm/checksum_64.h b/arch/sparc/include/asm/checksum_64.h
index b3c8b9ded364..b8779a6a5911 100644
--- a/arch/sparc/include/asm/checksum_64.h
+++ b/arch/sparc/include/asm/checksum_64.h
@@ -164,4 +164,16 @@ static inline __sum16 ip_compute_csum(const void *buff, int len)
 	return csum_fold(csum_partial(buff, len, 0));
 }
 
+#define HAVE_ARCH_CSUM_ADD
+static inline __wsum csum_add(__wsum csum, __wsum addend)
+{
+	__asm__ __volatile__(
+		"addcc   %0, %1, %0\n"
+		"addx    %0, %%g0, %0"
+		: "=r" (csum)
+		: "r" (addend), "0" (csum));
+
+	return csum;
+}
+
 #endif /* !(__SPARC64_CHECKSUM_H) */
diff --git a/arch/sparc/include/asm/pci_32.h b/arch/sparc/include/asm/pci_32.h
index dc503297481f..53e9b4987db0 100644
--- a/arch/sparc/include/asm/pci_32.h
+++ b/arch/sparc/include/asm/pci_32.h
@@ -16,11 +16,6 @@
 
 #define PCI_IRQ_NONE		0xffffffff
 
-static inline void pcibios_penalize_isa_irq(int irq, int active)
-{
-	/* We don't do dynamic PCI IRQ allocation */
-}
-
 /* Dynamic DMA mapping stuff.
  */
 #define PCI_DMA_BUS_IS_PHYS	(0)
diff --git a/arch/sparc/include/asm/pci_64.h b/arch/sparc/include/asm/pci_64.h
index 2c9a3e9292a4..bd00a6226169 100644
--- a/arch/sparc/include/asm/pci_64.h
+++ b/arch/sparc/include/asm/pci_64.h
@@ -16,11 +16,6 @@
 
 #define PCI_IRQ_NONE		0xffffffff
 
-static inline void pcibios_penalize_isa_irq(int irq, int active)
-{
-	/* We don't do dynamic PCI IRQ allocation */
-}
-
 /* The PCI address space does not equal the physical memory
  * address space.  The networking and block device layers use
  * this boolean for bounce buffer decisions.
diff --git a/arch/sparc/include/asm/pgtable_64.h b/arch/sparc/include/asm/pgtable_64.h
index 521b8d8c1f96..3770bf5c6e1b 100644
--- a/arch/sparc/include/asm/pgtable_64.h
+++ b/arch/sparc/include/asm/pgtable_64.h
@@ -24,7 +24,8 @@
 
 /* The kernel image occupies 0x4000000 to 0x6000000 (4MB --> 96MB).
  * The page copy blockops can use 0x6000000 to 0x8000000.
- * The TSB is mapped in the 0x8000000 to 0xa000000 range.
+ * The 8K TSB is mapped in the 0x8000000 to 0x8400000 range.
+ * The 4M TSB is mapped in the 0x8400000 to 0x8800000 range.
  * The PROM resides in an area spanning 0xf0000000 to 0x100000000.
  * The vmalloc area spans 0x100000000 to 0x200000000.
  * Since modules need to be in the lowest 32-bits of the address space,
@@ -33,7 +34,8 @@
  * 0x400000000.
  */
 #define	TLBTEMP_BASE		_AC(0x0000000006000000,UL)
-#define	TSBMAP_BASE		_AC(0x0000000008000000,UL)
+#define	TSBMAP_8K_BASE		_AC(0x0000000008000000,UL)
+#define	TSBMAP_4M_BASE		_AC(0x0000000008400000,UL)
 #define MODULES_VADDR		_AC(0x0000000010000000,UL)
 #define MODULES_LEN		_AC(0x00000000e0000000,UL)
 #define MODULES_END		_AC(0x00000000f0000000,UL)
@@ -71,6 +73,23 @@
 
 #include <linux/sched.h>
 
+extern unsigned long sparc64_valid_addr_bitmap[];
+
+/* Needs to be defined here and not in linux/mm.h, as it is arch dependent */
+static inline bool __kern_addr_valid(unsigned long paddr)
+{
+	if ((paddr >> MAX_PHYS_ADDRESS_BITS) != 0UL)
+		return false;
+	return test_bit(paddr >> ILOG2_4MB, sparc64_valid_addr_bitmap);
+}
+
+static inline bool kern_addr_valid(unsigned long addr)
+{
+	unsigned long paddr = __pa(addr);
+
+	return __kern_addr_valid(paddr);
+}
+
 /* Entries per page directory level. */
 #define PTRS_PER_PTE	(1UL << (PAGE_SHIFT-3))
 #define PTRS_PER_PMD	(1UL << PMD_BITS)
@@ -79,9 +98,12 @@
 /* Kernel has a separate 44bit address space. */
 #define FIRST_USER_ADDRESS	0
 
-#define pte_ERROR(e)	__builtin_trap()
-#define pmd_ERROR(e)	__builtin_trap()
-#define pgd_ERROR(e)	__builtin_trap()
+#define pmd_ERROR(e)							\
+	pr_err("%s:%d: bad pmd %p(%016lx) seen at (%pS)\n",		\
+	       __FILE__, __LINE__, &(e), pmd_val(e), __builtin_return_address(0))
+#define pgd_ERROR(e)							\
+	pr_err("%s:%d: bad pgd %p(%016lx) seen at (%pS)\n",		\
+	       __FILE__, __LINE__, &(e), pgd_val(e), __builtin_return_address(0))
 
 #endif /* !(__ASSEMBLY__) */
 
@@ -258,8 +280,8 @@ static inline pte_t pte_modify(pte_t pte, pgprot_t prot)
 {
 	unsigned long mask, tmp;
 
-	/* SUN4U: 0x600307ffffffecb8 (negated == 0x9ffcf80000001347)
-	 * SUN4V: 0x30ffffffffffee17 (negated == 0xcf000000000011e8)
+	/* SUN4U: 0x630107ffffffec38 (negated == 0x9cfef800000013c7)
+	 * SUN4V: 0x33ffffffffffee07 (negated == 0xcc000000000011f8)
 	 *
 	 * Even if we use negation tricks the result is still a 6
 	 * instruction sequence, so don't try to play fancy and just
@@ -289,10 +311,10 @@ static inline pte_t pte_modify(pte_t pte, pgprot_t prot)
 	"	.previous\n"
 	: "=r" (mask), "=r" (tmp)
 	: "i" (_PAGE_PADDR_4U | _PAGE_MODIFIED_4U | _PAGE_ACCESSED_4U |
-	       _PAGE_CP_4U | _PAGE_CV_4U | _PAGE_E_4U | _PAGE_PRESENT_4U |
+	       _PAGE_CP_4U | _PAGE_CV_4U | _PAGE_E_4U |
 	       _PAGE_SPECIAL | _PAGE_PMD_HUGE | _PAGE_SZALL_4U),
 	  "i" (_PAGE_PADDR_4V | _PAGE_MODIFIED_4V | _PAGE_ACCESSED_4V |
-	       _PAGE_CP_4V | _PAGE_CV_4V | _PAGE_E_4V | _PAGE_PRESENT_4V |
+	       _PAGE_CP_4V | _PAGE_CV_4V | _PAGE_E_4V |
 	       _PAGE_SPECIAL | _PAGE_PMD_HUGE | _PAGE_SZALL_4V));
 
 	return __pte((pte_val(pte) & mask) | (pgprot_val(prot) & ~mask));
@@ -633,7 +655,7 @@ static inline unsigned long pmd_large(pmd_t pmd)
 {
 	pte_t pte = __pte(pmd_val(pmd));
 
-	return (pte_val(pte) & _PAGE_PMD_HUGE) && pte_present(pte);
+	return pte_val(pte) & _PAGE_PMD_HUGE;
 }
 
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
@@ -719,20 +741,6 @@ static inline pmd_t pmd_mkwrite(pmd_t pmd)
 	return __pmd(pte_val(pte));
 }
 
-static inline pmd_t pmd_mknotpresent(pmd_t pmd)
-{
-	unsigned long mask;
-
-	if (tlb_type == hypervisor)
-		mask = _PAGE_PRESENT_4V;
-	else
-		mask = _PAGE_PRESENT_4U;
-
-	pmd_val(pmd) &= ~mask;
-
-	return pmd;
-}
-
 static inline pmd_t pmd_mksplitting(pmd_t pmd)
 {
 	pte_t pte = __pte(pmd_val(pmd));
@@ -757,6 +765,20 @@ static inline int pmd_present(pmd_t pmd)
 
 #define pmd_none(pmd)			(!pmd_val(pmd))
 
+/* pmd_bad() is only called on non-trans-huge PMDs.  Our encoding is
+ * very simple, it's just the physical address.  PTE tables are of
+ * size PAGE_SIZE so make sure the sub-PAGE_SIZE bits are clear and
+ * the top bits outside of the range of any physical address size we
+ * support are clear as well.  We also validate the physical itself.
+ */
+#define pmd_bad(pmd)			((pmd_val(pmd) & ~PAGE_MASK) || \
+					 !__kern_addr_valid(pmd_val(pmd)))
+
+#define pud_none(pud)			(!pud_val(pud))
+
+#define pud_bad(pud)			((pud_val(pud) & ~PAGE_MASK) || \
+					 !__kern_addr_valid(pud_val(pud)))
+
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
 void set_pmd_at(struct mm_struct *mm, unsigned long addr,
 		pmd_t *pmdp, pmd_t pmd);
@@ -790,10 +812,7 @@ static inline unsigned long __pmd_page(pmd_t pmd)
 #define pud_page_vaddr(pud)		\
 	((unsigned long) __va(pud_val(pud)))
 #define pud_page(pud) 			virt_to_page((void *)pud_page_vaddr(pud))
-#define pmd_bad(pmd)			(0)
 #define pmd_clear(pmdp)			(pmd_val(*(pmdp)) = 0UL)
-#define pud_none(pud)			(!pud_val(pud))
-#define pud_bad(pud)			(0)
 #define pud_present(pud)		(pud_val(pud) != 0U)
 #define pud_clear(pudp)			(pud_val(*(pudp)) = 0UL)
 
@@ -893,6 +912,10 @@ void update_mmu_cache(struct vm_area_struct *, unsigned long, pte_t *);
 void update_mmu_cache_pmd(struct vm_area_struct *vma, unsigned long addr,
 			  pmd_t *pmd);
 
+#define __HAVE_ARCH_PMDP_INVALIDATE
+extern void pmdp_invalidate(struct vm_area_struct *vma, unsigned long address,
+			    pmd_t *pmdp);
+
 #define __HAVE_ARCH_PGTABLE_DEPOSIT
 void pgtable_trans_huge_deposit(struct mm_struct *mm, pmd_t *pmdp,
 				pgtable_t pgtable);
@@ -919,18 +942,6 @@ unsigned long pte_file(pte_t);
 pte_t pgoff_to_pte(unsigned long);
 #define PTE_FILE_MAX_BITS	(64UL - PAGE_SHIFT - 1UL)
 
-extern unsigned long sparc64_valid_addr_bitmap[];
-
-/* Needs to be defined here and not in linux/mm.h, as it is arch dependent */
-static inline bool kern_addr_valid(unsigned long addr)
-{
-	unsigned long paddr = __pa(addr);
-
-	if ((paddr >> 41UL) != 0UL)
-		return false;
-	return test_bit(paddr >> 22, sparc64_valid_addr_bitmap);
-}
-
 int page_in_phys_avail(unsigned long paddr);
 
 /*
diff --git a/arch/sparc/include/asm/tsb.h b/arch/sparc/include/asm/tsb.h
index 2230f80d9fe3..90916f955cac 100644
--- a/arch/sparc/include/asm/tsb.h
+++ b/arch/sparc/include/asm/tsb.h
@@ -171,7 +171,8 @@ extern struct tsb_phys_patch_entry __tsb_phys_patch, __tsb_phys_patch_end;
 	andcc		REG1, REG2, %g0;		\
 	be,pt		%xcc, 700f;			\
 	 sethi		%hi(4 * 1024 * 1024), REG2;	\
-	andn		REG1, REG2, REG1;		\
+	brgez,pn	REG1, FAIL_LABEL;		\
+	 andn		REG1, REG2, REG1;		\
 	and		VADDR, REG2, REG2;		\
 	brlz,pt		REG1, PTE_LABEL;		\
 	 or		REG1, REG2, REG1;		\
diff --git a/arch/sparc/include/asm/unistd.h b/arch/sparc/include/asm/unistd.h
index dfa53fdd5cbc..0aac1e8f2968 100644
--- a/arch/sparc/include/asm/unistd.h
+++ b/arch/sparc/include/asm/unistd.h
@@ -25,7 +25,6 @@
 #define __ARCH_WANT_SYS_ALARM
 #define __ARCH_WANT_SYS_GETHOSTNAME
 #define __ARCH_WANT_SYS_PAUSE
-#define __ARCH_WANT_SYS_SGETMASK
 #define __ARCH_WANT_SYS_SIGNAL
 #define __ARCH_WANT_SYS_TIME
 #define __ARCH_WANT_SYS_UTIME