Merge branch 'master' into percpu

Conflicts:
	arch/powerpc/platforms/pseries/hvCall.S
	include/linux/percpu.h

54 files changed, 340 insertions, 502 deletions

diff --git a/arch/ia64/Makefile b/arch/ia64/Makefile
index e7cbaa02cd0..475e2725fbd 100644
--- a/arch/ia64/Makefile
+++ b/arch/ia64/Makefile
@@ -103,4 +103,4 @@ archprepare: make_nr_irqs_h FORCE
 PHONY += make_nr_irqs_h FORCE
 
 make_nr_irqs_h: FORCE
-	$(Q)$(MAKE) $(build)=arch/ia64/kernel include/asm-ia64/nr-irqs.h
+	$(Q)$(MAKE) $(build)=arch/ia64/kernel include/generated/nr-irqs.h
diff --git a/arch/ia64/hp/common/sba_iommu.c b/arch/ia64/hp/common/sba_iommu.c
index 674a8374c6d..e14c492a8a9 100644
--- a/arch/ia64/hp/common/sba_iommu.c
+++ b/arch/ia64/hp/common/sba_iommu.c
@@ -677,12 +677,19 @@ sba_alloc_range(struct ioc *ioc, struct device *dev, size_t size)
 			spin_unlock_irqrestore(&ioc->saved_lock, flags);
 
 			pide = sba_search_bitmap(ioc, dev, pages_needed, 0);
-			if (unlikely(pide >= (ioc->res_size << 3)))
-				panic(__FILE__ ": I/O MMU @ %p is out of mapping resources\n",
-				      ioc->ioc_hpa);
+			if (unlikely(pide >= (ioc->res_size << 3))) {
+				printk(KERN_WARNING "%s: I/O MMU @ %p is"
+				       "out of mapping resources, %u %u %lx\n",
+				       __func__, ioc->ioc_hpa, ioc->res_size,
+				       pages_needed, dma_get_seg_boundary(dev));
+				return -1;
+			}
 #else
-			panic(__FILE__ ": I/O MMU @ %p is out of mapping resources\n",
-			      ioc->ioc_hpa);
+			printk(KERN_WARNING "%s: I/O MMU @ %p is"
+			       "out of mapping resources, %u %u %lx\n",
+			       __func__, ioc->ioc_hpa, ioc->res_size,
+			       pages_needed, dma_get_seg_boundary(dev));
+			return -1;
 #endif
 		}
 	}
@@ -965,6 +972,8 @@ static dma_addr_t sba_map_page(struct device *dev, struct page *page,
 #endif
 
 	pide = sba_alloc_range(ioc, dev, size);
+	if (pide < 0)
+		return 0;
 
 	iovp = (dma_addr_t) pide << iovp_shift;
 
@@ -1320,6 +1329,7 @@ sba_coalesce_chunks(struct ioc *ioc, struct device *dev,
 	unsigned long dma_offset, dma_len; /* start/len of DMA stream */
 	int n_mappings = 0;
 	unsigned int max_seg_size = dma_get_max_seg_size(dev);
+	int idx;
 
 	while (nents > 0) {
 		unsigned long vaddr = (unsigned long) sba_sg_address(startsg);
@@ -1381,7 +1391,7 @@ sba_coalesce_chunks(struct ioc *ioc, struct device *dev,
 #endif
 
 			/*
-			** Not virtually contigous.
+			** Not virtually contiguous.
 			** Terminate prev chunk.
 			** Start a new chunk.
 			**
@@ -1418,16 +1428,22 @@ sba_coalesce_chunks(struct ioc *ioc, struct device *dev,
 		vcontig_sg->dma_length = vcontig_len;
 		dma_len = (dma_len + dma_offset + ~iovp_mask) & iovp_mask;
 		ASSERT(dma_len <= DMA_CHUNK_SIZE);
-		dma_sg->dma_address = (dma_addr_t) (PIDE_FLAG
-			| (sba_alloc_range(ioc, dev, dma_len) << iovp_shift)
-			| dma_offset);
+		idx = sba_alloc_range(ioc, dev, dma_len);
+		if (idx < 0) {
+			dma_sg->dma_length = 0;
+			return -1;
+		}
+		dma_sg->dma_address = (dma_addr_t)(PIDE_FLAG | (idx << iovp_shift)
+						   | dma_offset);
 		n_mappings++;
 	}
 
 	return n_mappings;
 }
 
-
+static void sba_unmap_sg_attrs(struct device *dev, struct scatterlist *sglist,
+			       int nents, enum dma_data_direction dir,
+			       struct dma_attrs *attrs);
 /**
  * sba_map_sg - map Scatter/Gather list
  * @dev: instance of PCI owned by the driver that's asking.
@@ -1493,6 +1509,10 @@ static int sba_map_sg_attrs(struct device *dev, struct scatterlist *sglist,
 	** Access to the virtual address is what forces a two pass algorithm.
 	*/
 	coalesced = sba_coalesce_chunks(ioc, dev, sglist, nents);
+	if (coalesced < 0) {
+		sba_unmap_sg_attrs(dev, sglist, nents, dir, attrs);
+		return 0;
+	}
 
 	/*
 	** Program the I/O Pdir
diff --git a/arch/ia64/ia32/elfcore32.h b/arch/ia64/ia32/elfcore32.h
index 9a3abf58cea..65772574261 100644
--- a/arch/ia64/ia32/elfcore32.h
+++ b/arch/ia64/ia32/elfcore32.h
@@ -11,8 +11,6 @@
 #include <asm/intrinsics.h>
 #include <asm/uaccess.h>
 
-#define USE_ELF_CORE_DUMP 1
-
 /* Override elfcore.h */
 #define _LINUX_ELFCORE_H 1
 typedef unsigned int elf_greg_t;
diff --git a/arch/ia64/ia32/ia32_entry.S b/arch/ia64/ia32/ia32_entry.S
index af9405cd70e..2fd7479aa21 100644
--- a/arch/ia64/ia32/ia32_entry.S
+++ b/arch/ia64/ia32/ia32_entry.S
@@ -79,7 +79,7 @@ GLOBAL_ENTRY(ia32_ret_from_clone)
 (p6)	br.cond.spnt .ia32_strace_check_retval
 	;;					// prevent RAW on r8
 END(ia32_ret_from_clone)
-	// fall thrugh
+	// fall through
 GLOBAL_ENTRY(ia32_ret_from_syscall)
 	PT_REGS_UNWIND_INFO(0)
 
@@ -327,7 +327,7 @@ ia32_syscall_table:
 	data8 compat_sys_writev
 	data8 sys_getsid
 	data8 sys_fdatasync
-	data8 sys32_sysctl
+	data8 compat_sys_sysctl
 	data8 sys_mlock		  /* 150 */
 	data8 sys_munlock
 	data8 sys_mlockall
diff --git a/arch/ia64/ia32/sys_ia32.c b/arch/ia64/ia32/sys_ia32.c
index 625ed8f76fc..045b746b980 100644
--- a/arch/ia64/ia32/sys_ia32.c
+++ b/arch/ia64/ia32/sys_ia32.c
@@ -858,6 +858,9 @@ ia32_do_mmap (struct file *file, unsigned long addr, unsigned long len, int prot
 
 	prot = get_prot32(prot);
 
+	if (flags & MAP_HUGETLB)
+		return -ENOMEM;
+
 #if PAGE_SHIFT > IA32_PAGE_SHIFT
 	mutex_lock(&ia32_mmap_mutex);
 	{
@@ -1628,61 +1631,6 @@ sys32_msync (unsigned int start, unsigned int len, int flags)
 	return sys_msync(addr, len + (start - addr), flags);
 }
 
-struct sysctl32 {
-	unsigned int	name;
-	int		nlen;
-	unsigned int	oldval;
-	unsigned int	oldlenp;
-	unsigned int	newval;
-	unsigned int	newlen;
-	unsigned int	__unused[4];
-};
-
-#ifdef CONFIG_SYSCTL_SYSCALL
-asmlinkage long
-sys32_sysctl (struct sysctl32 __user *args)
-{
-	struct sysctl32 a32;
-	mm_segment_t old_fs = get_fs ();
-	void __user *oldvalp, *newvalp;
-	size_t oldlen;
-	int __user *namep;
-	long ret;
-
-	if (copy_from_user(&a32, args, sizeof(a32)))
-		return -EFAULT;
-
-	/*
-	 * We need to pre-validate these because we have to disable address checking
-	 * before calling do_sysctl() because of OLDLEN but we can't run the risk of the
-	 * user specifying bad addresses here.  Well, since we're dealing with 32 bit
-	 * addresses, we KNOW that access_ok() will always succeed, so this is an
-	 * expensive NOP, but so what...
-	 */
-	namep = (int __user *) compat_ptr(a32.name);
-	oldvalp = compat_ptr(a32.oldval);
-	newvalp = compat_ptr(a32.newval);
-
-	if ((oldvalp && get_user(oldlen, (int __user *) compat_ptr(a32.oldlenp)))
-	    || !access_ok(VERIFY_WRITE, namep, 0)
-	    || !access_ok(VERIFY_WRITE, oldvalp, 0)
-	    || !access_ok(VERIFY_WRITE, newvalp, 0))
-		return -EFAULT;
-
-	set_fs(KERNEL_DS);
-	lock_kernel();
-	ret = do_sysctl(namep, a32.nlen, oldvalp, (size_t __user *) &oldlen,
-			newvalp, (size_t) a32.newlen);
-	unlock_kernel();
-	set_fs(old_fs);
-
-	if (oldvalp && put_user (oldlen, (int __user *) compat_ptr(a32.oldlenp)))
-		return -EFAULT;
-
-	return ret;
-}
-#endif
-
 asmlinkage long
 sys32_newuname (struct new_utsname __user *name)
 {
diff --git a/arch/ia64/include/asm/acpi.h b/arch/ia64/include/asm/acpi.h
index 91df9686a0d..7ae58892ba8 100644
--- a/arch/ia64/include/asm/acpi.h
+++ b/arch/ia64/include/asm/acpi.h
@@ -132,6 +132,12 @@ extern int __devinitdata pxm_to_nid_map[MAX_PXM_DOMAINS];
 extern int __initdata nid_to_pxm_map[MAX_NUMNODES];
 #endif
 
+static inline bool arch_has_acpi_pdc(void) { return true; }
+static inline void arch_acpi_set_pdc_bits(u32 *buf)
+{
+	buf[2] |= ACPI_PDC_EST_CAPABILITY_SMP;
+}
+
 #define acpi_unlazy_tlb(x)
 
 #ifdef CONFIG_ACPI_NUMA
diff --git a/arch/ia64/include/asm/asm-offsets.h b/arch/ia64/include/asm/asm-offsets.h
new file mode 100644
index 00000000000..d370ee36a18
--- /dev/null
+++ b/arch/ia64/include/asm/asm-offsets.h
@@ -0,0 +1 @@
+#include <generated/asm-offsets.h>
diff --git a/arch/ia64/include/asm/bitops.h b/arch/ia64/include/asm/bitops.h
index 57a2787bc9f..6ebc229a1c5 100644
--- a/arch/ia64/include/asm/bitops.h
+++ b/arch/ia64/include/asm/bitops.h
@@ -127,7 +127,7 @@ clear_bit_unlock (int nr, volatile void *addr)
  * @addr: Address to start counting from
  *
  * Similarly to clear_bit_unlock, the implementation uses a store
- * with release semantics. See also __raw_spin_unlock().
+ * with release semantics. See also arch_spin_unlock().
  */
 static __inline__ void
 __clear_bit_unlock(int nr, void *addr)
diff --git a/arch/ia64/include/asm/cacheflush.h b/arch/ia64/include/asm/cacheflush.h
index c8ce2719fee..429eefc93ee 100644
--- a/arch/ia64/include/asm/cacheflush.h
+++ b/arch/ia64/include/asm/cacheflush.h
@@ -25,6 +25,7 @@
 #define flush_cache_vmap(start, end)		do { } while (0)
 #define flush_cache_vunmap(start, end)		do { } while (0)
 
+#define ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE 1
 #define flush_dcache_page(page)			\
 do {						\
 	clear_bit(PG_arch_1, &(page)->flags);	\
diff --git a/arch/ia64/include/asm/dma-mapping.h b/arch/ia64/include/asm/dma-mapping.h
index 8d3c79cd81e..7d09a09cdaa 100644
--- a/arch/ia64/include/asm/dma-mapping.h
+++ b/arch/ia64/include/asm/dma-mapping.h
@@ -73,7 +73,7 @@ static inline bool dma_capable(struct device *dev, dma_addr_t addr, size_t size)
 	if (!dev->dma_mask)
 		return 0;
 
-	return addr + size <= *dev->dma_mask;
+	return addr + size - 1 <= *dev->dma_mask;
 }
 
 static inline dma_addr_t phys_to_dma(struct device *dev, phys_addr_t paddr)
diff --git a/arch/ia64/include/asm/elf.h b/arch/ia64/include/asm/elf.h
index 86eddee029c..e14108b19c0 100644
--- a/arch/ia64/include/asm/elf.h
+++ b/arch/ia64/include/asm/elf.h
@@ -25,7 +25,6 @@
 #define ELF_DATA	ELFDATA2LSB
 #define ELF_ARCH	EM_IA_64
 
-#define USE_ELF_CORE_DUMP
 #define CORE_DUMP_USE_REGSET
 
 /* Least-significant four bits of ELF header's e_flags are OS-specific.  The bits are
diff --git a/arch/ia64/include/asm/hw_irq.h b/arch/ia64/include/asm/hw_irq.h
index 91619b31dbf..bf2e37493e0 100644
--- a/arch/ia64/include/asm/hw_irq.h
+++ b/arch/ia64/include/asm/hw_irq.h
@@ -59,7 +59,13 @@ typedef u16 ia64_vector;
 extern int ia64_first_device_vector;
 extern int ia64_last_device_vector;
 
+#if defined(CONFIG_SMP) && (defined(CONFIG_IA64_GENERIC) || defined (CONFIG_IA64_DIG))
+/* Reserve the lower priority vector than device vectors for "move IRQ" IPI */
+#define IA64_IRQ_MOVE_VECTOR		0x30	/* "move IRQ" IPI */
+#define IA64_DEF_FIRST_DEVICE_VECTOR	0x31
+#else
 #define IA64_DEF_FIRST_DEVICE_VECTOR	0x30
+#endif
 #define IA64_DEF_LAST_DEVICE_VECTOR	0xe7
 #define IA64_FIRST_DEVICE_VECTOR	ia64_first_device_vector
 #define IA64_LAST_DEVICE_VECTOR		ia64_last_device_vector
diff --git a/arch/ia64/include/asm/io.h b/arch/ia64/include/asm/io.h
index 0d9d16e2d94..cc8335eb311 100644
--- a/arch/ia64/include/asm/io.h
+++ b/arch/ia64/include/asm/io.h
@@ -424,6 +424,8 @@ __writeq (unsigned long val, volatile void __iomem *addr)
 extern void __iomem * ioremap(unsigned long offset, unsigned long size);
 extern void __iomem * ioremap_nocache (unsigned long offset, unsigned long size);
 extern void iounmap (volatile void __iomem *addr);
+extern void __iomem * early_ioremap (unsigned long phys_addr, unsigned long size);
+extern void early_iounmap (volatile void __iomem *addr, unsigned long size);
 
 /*
  * String version of IO memory access ops:
diff --git a/arch/ia64/include/asm/irq.h b/arch/ia64/include/asm/irq.h
index 5282546cdf8..91b920fd7d5 100644
--- a/arch/ia64/include/asm/irq.h
+++ b/arch/ia64/include/asm/irq.h
@@ -13,7 +13,7 @@
 
 #include <linux/types.h>
 #include <linux/cpumask.h>
-#include <asm-ia64/nr-irqs.h>
+#include <generated/nr-irqs.h>
 
 static __inline__ int
 irq_canonicalize (int irq)
diff --git a/arch/ia64/include/asm/kvm.h b/arch/ia64/include/asm/kvm.h
index 18a7e49abbc..bc90c75adf6 100644
--- a/arch/ia64/include/asm/kvm.h
+++ b/arch/ia64/include/asm/kvm.h
@@ -60,6 +60,7 @@ struct kvm_ioapic_state {
 #define KVM_IRQCHIP_PIC_MASTER   0
 #define KVM_IRQCHIP_PIC_SLAVE    1
 #define KVM_IRQCHIP_IOAPIC       2
+#define KVM_NR_IRQCHIPS          3
 
 #define KVM_CONTEXT_SIZE	8*1024
 
diff --git a/arch/ia64/include/asm/kvm_host.h b/arch/ia64/include/asm/kvm_host.h
index d9b6325a932..a362e67e0ca 100644
--- a/arch/ia64/include/asm/kvm_host.h
+++ b/arch/ia64/include/asm/kvm_host.h
@@ -475,7 +475,6 @@ struct kvm_arch {
 	struct list_head assigned_dev_head;
 	struct iommu_domain *iommu_domain;
 	int iommu_flags;
-	struct hlist_head irq_ack_notifier_list;
 
 	unsigned long irq_sources_bitmap;
 	unsigned long irq_states[KVM_IOAPIC_NUM_PINS];
diff --git a/arch/ia64/include/asm/mca.h b/arch/ia64/include/asm/mca.h
index c171cdf0a78..43f96ab18fa 100644
--- a/arch/ia64/include/asm/mca.h
+++ b/arch/ia64/include/asm/mca.h
@@ -106,6 +106,11 @@ struct ia64_sal_os_state {
 	unsigned long		os_status;		/* OS status to SAL, enum below */
 	unsigned long		context;		/* 0 if return to same context
 							   1 if return to new context */
+
+	/* I-resources */
+	unsigned long		iip;
+	unsigned long		ipsr;
+	unsigned long		ifs;
 };
 
 enum {
diff --git a/arch/ia64/include/asm/numa.h b/arch/ia64/include/asm/numa.h
index 3499ff57bf4..6a8a27cfae3 100644
--- a/arch/ia64/include/asm/numa.h
+++ b/arch/ia64/include/asm/numa.h
@@ -22,8 +22,6 @@
 
 #include <asm/mmzone.h>
 
-#define NUMA_NO_NODE	-1
-
 extern u16 cpu_to_node_map[NR_CPUS] __cacheline_aligned;
 extern cpumask_t node_to_cpu_mask[MAX_NUMNODES] __cacheline_aligned;
 extern pg_data_t *pgdat_list[MAX_NUMNODES];
diff --git a/arch/ia64/include/asm/perfmon_default_smpl.h b/arch/ia64/include/asm/perfmon_default_smpl.h
index 48822c0811d..74724b24c2b 100644
--- a/arch/ia64/include/asm/perfmon_default_smpl.h
+++ b/arch/ia64/include/asm/perfmon_default_smpl.h
@@ -67,7 +67,7 @@ typedef struct {
         unsigned long   ip;                     /* where did the overflow interrupt happened  */
         unsigned long   tstamp;                 /* ar.itc when entering perfmon intr. handler */
 
-        unsigned short  cpu;                    /* cpu on which the overfow occured */
+        unsigned short  cpu;                    /* cpu on which the overflow occured */
         unsigned short  set;                    /* event set active when overflow ocurred   */
         int    		tgid;              	/* thread group id (for NPTL, this is getpid()) */
 } pfm_default_smpl_entry_t;
diff --git a/arch/ia64/include/asm/rwsem.h b/arch/ia64/include/asm/rwsem.h
index fbee74b1578..e8762688e8e 100644
--- a/arch/ia64/include/asm/rwsem.h
+++ b/arch/ia64/include/asm/rwsem.h
@@ -47,7 +47,7 @@ struct rw_semaphore {
 #define RWSEM_ACTIVE_WRITE_BIAS		(RWSEM_WAITING_BIAS + RWSEM_ACTIVE_BIAS)
 
 #define __RWSEM_INITIALIZER(name) \
-	{ RWSEM_UNLOCKED_VALUE, SPIN_LOCK_UNLOCKED, \
+	{ RWSEM_UNLOCKED_VALUE, __SPIN_LOCK_UNLOCKED((name).wait_lock), \
 	  LIST_HEAD_INIT((name).wait_list) }
 
 #define DECLARE_RWSEM(name) \
diff --git a/arch/ia64/include/asm/sn/shubio.h b/arch/ia64/include/asm/sn/shubio.h
index 22a6f18a531..6052422a22b 100644
--- a/arch/ia64/include/asm/sn/shubio.h
+++ b/arch/ia64/include/asm/sn/shubio.h
@@ -3289,7 +3289,7 @@ typedef ii_icrb0_e_u_t icrbe_t;
 #define IIO_IIDSR_LVL_SHIFT     0
 #define IIO_IIDSR_LVL_MASK      0x000000ff
 
-/* Xtalk timeout threshhold register (IIO_IXTT) */
+/* Xtalk timeout threshold register (IIO_IXTT) */
 #define IXTT_RRSP_TO_SHFT	55	/* read response timeout */
 #define IXTT_RRSP_TO_MASK	(0x1FULL << IXTT_RRSP_TO_SHFT)
 #define IXTT_RRSP_PS_SHFT	32	/* read responsed TO prescalar */
diff --git a/arch/ia64/include/asm/socket.h b/arch/ia64/include/asm/socket.h
index 0b0d5ff062e..51427eaa51b 100644
--- a/arch/ia64/include/asm/socket.h
+++ b/arch/ia64/include/asm/socket.h
@@ -69,4 +69,6 @@
 #define SO_PROTOCOL		38
 #define SO_DOMAIN		39
 
+#define SO_RXQ_OVFL             40
+
 #endif /* _ASM_IA64_SOCKET_H */
diff --git a/arch/ia64/include/asm/spinlock.h b/arch/ia64/include/asm/spinlock.h
index 30bb930e111..1a91c9121d1 100644
--- a/arch/ia64/include/asm/spinlock.h
+++ b/arch/ia64/include/asm/spinlock.h
@@ -17,7 +17,7 @@
 #include <asm/intrinsics.h>
 #include <asm/system.h>
 
-#define __raw_spin_lock_init(x)			((x)->lock = 0)
+#define arch_spin_lock_init(x)			((x)->lock = 0)
 
 /*
  * Ticket locks are conceptually two parts, one indicating the current head of
@@ -25,108 +25,128 @@
  * by atomically noting the tail and incrementing it by one (thus adding
  * ourself to the queue and noting our position), then waiting until the head
  * becomes equal to the the initial value of the tail.
+ * The pad bits in the middle are used to prevent the next_ticket number
+ * overflowing into the now_serving number.
  *
- *   63                     32  31                      0
+ *   31             17  16    15  14                    0
  *  +----------------------------------------------------+
- *  |  next_ticket_number      |     now_serving         |
+ *  |  now_serving     | padding |   next_ticket         |
  *  +----------------------------------------------------+
  */
 
-#define TICKET_SHIFT	32
+#define TICKET_SHIFT	17
+#define TICKET_BITS	15
+#define	TICKET_MASK	((1 << TICKET_BITS) - 1)
 
-static __always_inline void __ticket_spin_lock(raw_spinlock_t *lock)
+static __always_inline void __ticket_spin_lock(arch_spinlock_t *lock)
 {
-	int	*p = (int *)&lock->lock, turn, now_serving;
+	int	*p = (int *)&lock->lock, ticket, serve;
 
-	now_serving = *p;
-	turn = ia64_fetchadd(1, p+1, acq);
+	ticket = ia64_fetchadd(1, p, acq);
 
-	if (turn == now_serving)
+	if (!(((ticket >> TICKET_SHIFT) ^ ticket) & TICKET_MASK))
 		return;
 
-	do {
+	ia64_invala();
+
+	for (;;) {
+		asm volatile ("ld4.c.nc %0=[%1]" : "=r"(serve) : "r"(p) : "memory");
+
+		if (!(((serve >> TICKET_SHIFT) ^ ticket) & TICKET_MASK))
+			return;
 		cpu_relax();
-	} while (ACCESS_ONCE(*p) != turn);
+	}
 }
 
-static __always_inline int __ticket_spin_trylock(raw_spinlock_t *lock)
+static __always_inline int __ticket_spin_trylock(arch_spinlock_t *lock)
 {
-	long tmp = ACCESS_ONCE(lock->lock), try;
-
-	if (!(((tmp >> TICKET_SHIFT) ^ tmp) & ((1L << TICKET_SHIFT) - 1))) {
-		try = tmp + (1L << TICKET_SHIFT);
+	int tmp = ACCESS_ONCE(lock->lock);
 
-		return ia64_cmpxchg(acq, &lock->lock, tmp, try, sizeof (tmp)) == tmp;
-	}
+	if (!(((tmp >> TICKET_SHIFT) ^ tmp) & TICKET_MASK))
+		return ia64_cmpxchg(acq, &lock->lock, tmp, tmp + 1, sizeof (tmp)) == tmp;
 	return 0;
 }
 
-static __always_inline void __ticket_spin_unlock(raw_spinlock_t *lock)
+static __always_inline void __ticket_spin_unlock(arch_spinlock_t *lock)
 {
-	int	*p = (int *)&lock->lock;
+	unsigned short	*p = (unsigned short *)&lock->lock + 1, tmp;
 
-	(void)ia64_fetchadd(1, p, rel);
+	asm volatile ("ld2.bias %0=[%1]" : "=r"(tmp) : "r"(p));
+	ACCESS_ONCE(*p) = (tmp + 2) & ~1;
 }
 
-static inline int __ticket_spin_is_locked(raw_spinlock_t *lock)
+static __always_inline void __ticket_spin_unlock_wait(arch_spinlock_t *lock)
+{
+	int	*p = (int *)&lock->lock, ticket;
+
+	ia64_invala();
+
+	for (;;) {
+		asm volatile ("ld4.c.nc %0=[%1]" : "=r"(ticket) : "r"(p) : "memory");
+		if (!(((ticket >> TICKET_SHIFT) ^ ticket) & TICKET_MASK))
+			return;
+		cpu_relax();
+	}
+}
+
+static inline int __ticket_spin_is_locked(arch_spinlock_t *lock)
 {
 	long tmp = ACCESS_ONCE(lock->lock);
 
-	return !!(((tmp >> TICKET_SHIFT) ^ tmp) & ((1L << TICKET_SHIFT) - 1));
+	return !!(((tmp >> TICKET_SHIFT) ^ tmp) & TICKET_MASK);
 }
 
-static inline int __ticket_spin_is_contended(raw_spinlock_t *lock)
+static inline int __ticket_spin_is_contended(arch_spinlock_t *lock)
 {
 	long tmp = ACCESS_ONCE(lock->lock);
 
-	return (((tmp >> TICKET_SHIFT) - tmp) & ((1L << TICKET_SHIFT) - 1)) > 1;