Merge branch 'for-linus' of git://git390.osdl.marist.edu/pub/scm/linux-2.6

* 'for-linus' of git://git390.osdl.marist.edu/pub/scm/linux-2.6: [S390] 4level-fixup cleanup [S390] Cleanup page table definitions. [S390] Introduce follow_table in uaccess_pt.c [S390] Remove unused user_seg from thread structure. [S390] tlb flush fix. [S390] kernel: Fix dump on panic for DASDs under LPAR. [S390] struct class_device -> struct device conversion. [S390] cio: Fix incomplete commit for uevent suppression. [S390] cio: Use to_channelpath() for device to channel path conversion. [S390] Add per-cpu idle time / idle count sysfs attributes. [S390] Update default configuration.
author: Linus Torvalds <torvalds@woody.linux-foundation.org> 2007-10-22 19:23:34 -0700
committer: Linus Torvalds <torvalds@woody.linux-foundation.org> 2007-10-22 19:23:34 -0700
commit: 56d61a0e26c5a61c66d1ac259a59960295939da9 (patch)
tree: a23a30a966fe4220060682179294087cba1f9c57
parent: 5f48b338cd28f4095697a174d7e3e72084aca893 (diff)
parent: 190a1d722a59725706daf832bc8a511ed62f249d (diff)
23 files changed, 918 insertions, 713 deletions
diff --git a/arch/s390/defconfig b/arch/s390/defconfig
index 2aae23dba4b..ece7b99da89 100644
--- a/arch/s390/defconfig
+++ b/arch/s390/defconfig
@@ -1,7 +1,7 @@
 #
 # Automatically generated make config: don't edit
-# Linux kernel version: 2.6.22
-# Tue Jul 17 12:50:23 2007
+# Linux kernel version: 2.6.23
+# Mon Oct 22 12:10:44 2007
 #
 CONFIG_MMU=y
 CONFIG_ZONE_DMA=y
@@ -19,15 +19,11 @@ CONFIG_S390=y
 CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config"
 
 #
-# Code maturity level options
+# General setup
 #
 CONFIG_EXPERIMENTAL=y
 CONFIG_LOCK_KERNEL=y
 CONFIG_INIT_ENV_ARG_LIMIT=32
-
-#
-# General setup
-#
 CONFIG_LOCALVERSION=""
 CONFIG_LOCALVERSION_AUTO=y
 CONFIG_SWAP=y
@@ -42,7 +38,14 @@ CONFIG_AUDIT=y
 CONFIG_IKCONFIG=y
 CONFIG_IKCONFIG_PROC=y
 CONFIG_LOG_BUF_SHIFT=17
+CONFIG_CGROUPS=y
+# CONFIG_CGROUP_DEBUG is not set
+CONFIG_CGROUP_NS=y
+CONFIG_CGROUP_CPUACCT=y
 # CONFIG_CPUSETS is not set
+CONFIG_FAIR_GROUP_SCHED=y
+CONFIG_FAIR_USER_SCHED=y
+# CONFIG_FAIR_CGROUP_SCHED is not set
 CONFIG_SYSFS_DEPRECATED=y
 # CONFIG_RELAY is not set
 CONFIG_BLK_DEV_INITRD=y
@@ -63,7 +66,6 @@ CONFIG_FUTEX=y
 CONFIG_ANON_INODES=y
 CONFIG_EPOLL=y
 CONFIG_SIGNALFD=y
-CONFIG_TIMERFD=y
 CONFIG_EVENTFD=y
 CONFIG_SHMEM=y
 CONFIG_VM_EVENT_COUNTERS=y
@@ -83,6 +85,7 @@ CONFIG_STOP_MACHINE=y
 CONFIG_BLOCK=y
 # CONFIG_BLK_DEV_IO_TRACE is not set
 CONFIG_BLK_DEV_BSG=y
+CONFIG_BLOCK_COMPAT=y
 
 #
 # IO Schedulers
@@ -108,7 +111,6 @@ CONFIG_64BIT=y
 CONFIG_SMP=y
 CONFIG_NR_CPUS=32
 CONFIG_HOTPLUG_CPU=y
-CONFIG_DEFAULT_MIGRATION_COST=1000000
 CONFIG_COMPAT=y
 CONFIG_SYSVIPC_COMPAT=y
 CONFIG_AUDIT_ARCH=y
@@ -143,9 +145,11 @@ CONFIG_FLATMEM_MANUAL=y
 CONFIG_FLATMEM=y
 CONFIG_FLAT_NODE_MEM_MAP=y
 # CONFIG_SPARSEMEM_STATIC is not set
+# CONFIG_SPARSEMEM_VMEMMAP_ENABLE is not set
 CONFIG_SPLIT_PTLOCK_CPUS=4
 CONFIG_RESOURCES_64BIT=y
 CONFIG_ZONE_DMA_FLAG=1
+CONFIG_BOUNCE=y
 CONFIG_VIRT_TO_BUS=y
 CONFIG_HOLES_IN_ZONE=y
 
@@ -219,12 +223,14 @@ CONFIG_INET_TUNNEL=y
 CONFIG_INET_XFRM_MODE_TRANSPORT=y
 CONFIG_INET_XFRM_MODE_TUNNEL=y
 CONFIG_INET_XFRM_MODE_BEET=y
+CONFIG_INET_LRO=y
 CONFIG_INET_DIAG=y
 CONFIG_INET_TCP_DIAG=y
 # CONFIG_TCP_CONG_ADVANCED is not set
 CONFIG_TCP_CONG_CUBIC=y
 CONFIG_DEFAULT_TCP_CONG="cubic"
 # CONFIG_TCP_MD5SIG is not set
+# CONFIG_IP_VS is not set
 CONFIG_IPV6=y
 # CONFIG_IPV6_PRIVACY is not set
 # CONFIG_IPV6_ROUTER_PREF is not set
@@ -243,7 +249,48 @@ CONFIG_IPV6_SIT=y
 # CONFIG_IPV6_TUNNEL is not set
 # CONFIG_IPV6_MULTIPLE_TABLES is not set
 # CONFIG_NETWORK_SECMARK is not set
-# CONFIG_NETFILTER is not set
+CONFIG_NETFILTER=y
+# CONFIG_NETFILTER_DEBUG is not set
+
+#
+# Core Netfilter Configuration
+#
+CONFIG_NETFILTER_NETLINK=m
+CONFIG_NETFILTER_NETLINK_QUEUE=m
+CONFIG_NETFILTER_NETLINK_LOG=m
+CONFIG_NF_CONNTRACK_ENABLED=m
+CONFIG_NF_CONNTRACK=m
+# CONFIG_NF_CT_ACCT is not set
+# CONFIG_NF_CONNTRACK_MARK is not set
+# CONFIG_NF_CONNTRACK_EVENTS is not set
+# CONFIG_NF_CT_PROTO_SCTP is not set
+# CONFIG_NF_CT_PROTO_UDPLITE is not set
+# CONFIG_NF_CONNTRACK_AMANDA is not set
+# CONFIG_NF_CONNTRACK_FTP is not set
+# CONFIG_NF_CONNTRACK_H323 is not set
+# CONFIG_NF_CONNTRACK_IRC is not set
+# CONFIG_NF_CONNTRACK_NETBIOS_NS is not set
+# CONFIG_NF_CONNTRACK_PPTP is not set
+# CONFIG_NF_CONNTRACK_SANE is not set
+# CONFIG_NF_CONNTRACK_SIP is not set
+# CONFIG_NF_CONNTRACK_TFTP is not set
+# CONFIG_NF_CT_NETLINK is not set
+# CONFIG_NETFILTER_XTABLES is not set
+
+#
+# IP: Netfilter Configuration
+#
+# CONFIG_NF_CONNTRACK_IPV4 is not set
+# CONFIG_IP_NF_QUEUE is not set
+# CONFIG_IP_NF_IPTABLES is not set
+# CONFIG_IP_NF_ARPTABLES is not set
+
+#
+# IPv6: Netfilter Configuration (EXPERIMENTAL)
+#
+# CONFIG_NF_CONNTRACK_IPV6 is not set
+# CONFIG_IP6_NF_QUEUE is not set
+# CONFIG_IP6_NF_IPTABLES is not set
 # CONFIG_IP_DCCP is not set
 CONFIG_IP_SCTP=m
 # CONFIG_SCTP_DBG_MSG is not set
@@ -263,12 +310,7 @@ CONFIG_SCTP_HMAC_MD5=y
 # CONFIG_LAPB is not set
 # CONFIG_ECONET is not set
 # CONFIG_WAN_ROUTER is not set
-
-#
-# QoS and/or fair queueing
-#
 CONFIG_NET_SCHED=y
-CONFIG_NET_SCH_FIFO=y
 
 #
 # Queueing/Scheduling
@@ -306,10 +348,12 @@ CONFIG_NET_CLS_ACT=y
 CONFIG_NET_ACT_POLICE=y
 # CONFIG_NET_ACT_GACT is not set
 # CONFIG_NET_ACT_MIRRED is not set
+CONFIG_NET_ACT_NAT=m
 # CONFIG_NET_ACT_PEDIT is not set
 # CONFIG_NET_ACT_SIMP is not set
 CONFIG_NET_CLS_POLICE=y
 # CONFIG_NET_CLS_IND is not set
+CONFIG_NET_SCH_FIFO=y
 
 #
 # Network testing
@@ -329,6 +373,7 @@ CONFIG_CCW=y
 #
 # Generic Driver Options
 #
+CONFIG_UEVENT_HELPER_PATH="/sbin/hotplug"
 CONFIG_STANDALONE=y
 CONFIG_PREVENT_FIRMWARE_BUILD=y
 # CONFIG_FW_LOADER is not set
@@ -400,17 +445,11 @@ CONFIG_SCSI_FC_ATTRS=y
 # CONFIG_SCSI_ISCSI_ATTRS is not set
 # CONFIG_SCSI_SAS_ATTRS is not set
 # CONFIG_SCSI_SAS_LIBSAS is not set
-
-#
-# SCSI low-level drivers
-#
+# CONFIG_SCSI_SRP_ATTRS is not set
+CONFIG_SCSI_LOWLEVEL=y
 # CONFIG_ISCSI_TCP is not set
 # CONFIG_SCSI_DEBUG is not set
 CONFIG_ZFCP=y
-
-#
-# Multi-device support (RAID and LVM)
-#
 CONFIG_MD=y
 CONFIG_BLK_DEV_MD=y
 CONFIG_MD_LINEAR=m
@@ -429,7 +468,9 @@ CONFIG_DM_ZERO=y
 CONFIG_DM_MULTIPATH=y
 # CONFIG_DM_MULTIPATH_EMC is not set
 # CONFIG_DM_MULTIPATH_RDAC is not set
+# CONFIG_DM_MULTIPATH_HP is not set
 # CONFIG_DM_DELAY is not set
+# CONFIG_DM_UEVENT is not set
 CONFIG_NETDEVICES=y
 # CONFIG_NETDEVICES_MULTIQUEUE is not set
 # CONFIG_IFB is not set
@@ -438,8 +479,13 @@ CONFIG_BONDING=m
 # CONFIG_MACVLAN is not set
 CONFIG_EQUALIZER=m
 CONFIG_TUN=m
+CONFIG_VETH=m
 CONFIG_NET_ETHERNET=y
 # CONFIG_MII is not set
+# CONFIG_IBM_NEW_EMAC_ZMII is not set
+# CONFIG_IBM_NEW_EMAC_RGMII is not set
+# CONFIG_IBM_NEW_EMAC_TAH is not set
+# CONFIG_IBM_NEW_EMAC_EMAC4 is not set
 CONFIG_NETDEV_1000=y
 CONFIG_NETDEV_10000=y
 # CONFIG_TR is not set
@@ -473,7 +519,6 @@ CONFIG_CCWGROUP=y
 CONFIG_UNIX98_PTYS=y
 CONFIG_LEGACY_PTYS=y
 CONFIG_LEGACY_PTY_COUNT=256
-# CONFIG_WATCHDOG is not set
 CONFIG_HW_RANDOM=m
 # CONFIG_R3964 is not set
 CONFIG_RAW_DRIVER=m
@@ -490,7 +535,6 @@ CONFIG_TN3270_CONSOLE=y
 CONFIG_TN3215=y
 CONFIG_TN3215_CONSOLE=y
 CONFIG_CCW_CONSOLE=y
-CONFIG_SCLP=y
 CONFIG_SCLP_TTY=y
 CONFIG_SCLP_CONSOLE=y
 CONFIG_SCLP_VT220_TTY=y
@@ -514,6 +558,11 @@ CONFIG_S390_TAPE_34XX=m
 CONFIG_MONWRITER=m
 CONFIG_S390_VMUR=m
 # CONFIG_POWER_SUPPLY is not set
+# CONFIG_WATCHDOG is not set
+
+#
+# Sonics Silicon Backplane
+#
 
 #
 # File systems
@@ -569,7 +618,6 @@ CONFIG_SYSFS=y
 CONFIG_TMPFS=y
 CONFIG_TMPFS_POSIX_ACL=y
 # CONFIG_HUGETLB_PAGE is not set
-CONFIG_RAMFS=y
 CONFIG_CONFIGFS_FS=m
 
 #
@@ -588,10 +636,7 @@ CONFIG_CONFIGFS_FS=m
 # CONFIG_QNX4FS_FS is not set
 # CONFIG_SYSV_FS is not set
 # CONFIG_UFS_FS is not set
-
-#
-# Network File Systems
-#
+CONFIG_NETWORK_FILESYSTEMS=y
 CONFIG_NFS_FS=y
 CONFIG_NFS_V3=y
 # CONFIG_NFS_V3_ACL is not set
@@ -638,27 +683,13 @@ CONFIG_MSDOS_PARTITION=y
 # CONFIG_KARMA_PARTITION is not set
 # CONFIG_EFI_PARTITION is not set
 # CONFIG_SYSV68_PARTITION is not set
-
-#
-# Native Language Support
-#
 # CONFIG_NLS is not set
-
-#
-# Distributed Lock Manager
-#
 CONFIG_DLM=m
 # CONFIG_DLM_DEBUG is not set
-
-#
-# Instrumentation Support
-#
-
-#
-# Profiling support
-#
+CONFIG_INSTRUMENTATION=y
 # CONFIG_PROFILING is not set
 CONFIG_KPROBES=y
+# CONFIG_MARKERS is not set
 
 #
 # Kernel hacking
@@ -682,6 +713,7 @@ CONFIG_DEBUG_SPINLOCK=y
 CONFIG_DEBUG_MUTEXES=y
 # CONFIG_DEBUG_LOCK_ALLOC is not set
 # CONFIG_PROVE_LOCKING is not set
+# CONFIG_LOCK_STAT is not set
 CONFIG_DEBUG_SPINLOCK_SLEEP=y
 # CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set
 # CONFIG_DEBUG_KOBJECT is not set
@@ -694,14 +726,17 @@ CONFIG_FORCED_INLINING=y
 # CONFIG_RCU_TORTURE_TEST is not set
 # CONFIG_LKDTM is not set
 # CONFIG_FAULT_INJECTION is not set
+CONFIG_SAMPLES=y
 
 #
 # Security options
 #
 # CONFIG_KEYS is not set
 # CONFIG_SECURITY is not set
+# CONFIG_SECURITY_FILE_CAPABILITIES is not set
 CONFIG_CRYPTO=y
 CONFIG_CRYPTO_ALGAPI=y
+CONFIG_CRYPTO_AEAD=m
 CONFIG_CRYPTO_BLKCIPHER=y
 CONFIG_CRYPTO_HASH=m
 CONFIG_CRYPTO_MANAGER=y
@@ -720,6 +755,7 @@ CONFIG_CRYPTO_ECB=m
 CONFIG_CRYPTO_CBC=y
 CONFIG_CRYPTO_PCBC=m
 # CONFIG_CRYPTO_LRW is not set
+# CONFIG_CRYPTO_XTS is not set
 # CONFIG_CRYPTO_CRYPTD is not set
 # CONFIG_CRYPTO_DES is not set
 CONFIG_CRYPTO_FCRYPT=m
@@ -733,11 +769,13 @@ CONFIG_CRYPTO_FCRYPT=m
 # CONFIG_CRYPTO_ARC4 is not set
 # CONFIG_CRYPTO_KHAZAD is not set
 # CONFIG_CRYPTO_ANUBIS is not set
+CONFIG_CRYPTO_SEED=m
 # CONFIG_CRYPTO_DEFLATE is not set
 # CONFIG_CRYPTO_MICHAEL_MIC is not set
 # CONFIG_CRYPTO_CRC32C is not set
 CONFIG_CRYPTO_CAMELLIA=m
 # CONFIG_CRYPTO_TEST is not set
+CONFIG_CRYPTO_AUTHENC=m
 CONFIG_CRYPTO_HW=y
 # CONFIG_CRYPTO_SHA1_S390 is not set
 # CONFIG_CRYPTO_SHA256_S390 is not set
@@ -755,5 +793,6 @@ CONFIG_BITREVERSE=m
 # CONFIG_CRC16 is not set
 # CONFIG_CRC_ITU_T is not set
 CONFIG_CRC32=m
+CONFIG_CRC7=m
 # CONFIG_LIBCRC32C is not set
 CONFIG_PLIST=y
diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c
index 66b51901c87..ce0856d3250 100644
--- a/arch/s390/kernel/ipl.c
+++ b/arch/s390/kernel/ipl.c
@@ -648,6 +648,8 @@ static int dump_set_type(enum dump_type type)
 	case DUMP_TYPE_CCW:
 		if (MACHINE_IS_VM)
 			dump_method = DUMP_METHOD_CCW_VM;
+		else if (diag308_set_works)
+			dump_method = DUMP_METHOD_CCW_DIAG;
 		else
 			dump_method = DUMP_METHOD_CCW_CIO;
 		break;
diff --git a/arch/s390/kernel/process.c b/arch/s390/kernel/process.c
index 70c57378f42..96492cf2d49 100644
--- a/arch/s390/kernel/process.c
+++ b/arch/s390/kernel/process.c
@@ -44,6 +44,7 @@
 #include <asm/processor.h>
 #include <asm/irq.h>
 #include <asm/timer.h>
+#include <asm/cpu.h>
 
 asmlinkage void ret_from_fork(void) asm ("ret_from_fork");
 
@@ -91,6 +92,14 @@ EXPORT_SYMBOL(unregister_idle_notifier);
 
 void do_monitor_call(struct pt_regs *regs, long interruption_code)
 {
+	struct s390_idle_data *idle;
+
+	idle = &__get_cpu_var(s390_idle);
+	spin_lock(&idle->lock);
+	idle->idle_time += get_clock() - idle->idle_enter;
+	idle->in_idle = 0;
+	spin_unlock(&idle->lock);
+
 	/* disable monitor call class 0 */
 	__ctl_clear_bit(8, 15);
 
@@ -105,6 +114,7 @@ extern void s390_handle_mcck(void);
 static void default_idle(void)
 {
 	int cpu, rc;
+	struct s390_idle_data *idle;
 
 	/* CPU is going idle. */
 	cpu = smp_processor_id();
@@ -142,6 +152,12 @@ static void default_idle(void)
 		return;
 	}
 
+	idle = &__get_cpu_var(s390_idle);
+	spin_lock(&idle->lock);
+	idle->idle_count++;
+	idle->in_idle = 1;
+	idle->idle_enter = get_clock();
+	spin_unlock(&idle->lock);
 	trace_hardirqs_on();
 	/* Wait for external, I/O or machine check interrupt. */
 	__load_psw_mask(psw_kernel_bits | PSW_MASK_WAIT |
@@ -254,14 +270,12 @@ int copy_thread(int nr, unsigned long clone_flags, unsigned long new_stackp,
 	save_fp_regs(&current->thread.fp_regs);
 	memcpy(&p->thread.fp_regs, &current->thread.fp_regs,
 	       sizeof(s390_fp_regs));
-        p->thread.user_seg = __pa((unsigned long) p->mm->pgd) | _SEGMENT_TABLE;
 	/* Set a new TLS ?  */
 	if (clone_flags & CLONE_SETTLS)
 		p->thread.acrs[0] = regs->gprs[6];
 #else /* CONFIG_64BIT */
 	/* Save the fpu registers to new thread structure. */
 	save_fp_regs(&p->thread.fp_regs);
-        p->thread.user_seg = __pa((unsigned long) p->mm->pgd) | _REGION_TABLE;
 	/* Set a new TLS ?  */
 	if (clone_flags & CLONE_SETTLS) {
 		if (test_thread_flag(TIF_31BIT)) {
diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c
index 35edbef1d22..1d97fe1c0e5 100644
--- a/arch/s390/kernel/smp.c
+++ b/arch/s390/kernel/smp.c
@@ -42,6 +42,7 @@
 #include <asm/tlbflush.h>
 #include <asm/timer.h>
 #include <asm/lowcore.h>
+#include <asm/cpu.h>
 
 /*
  * An array with a pointer the lowcore of every CPU.
@@ -325,7 +326,7 @@ static void smp_ext_bitcall(int cpu, ec_bit_sig sig)
  */
 void smp_ptlb_callback(void *info)
 {
-	local_flush_tlb();
+	__tlb_flush_local();
 }
 
 void smp_ptlb_all(void)
@@ -494,6 +495,8 @@ int __cpuinit start_secondary(void *cpuvoid)
 	return 0;
 }
 
+DEFINE_PER_CPU(struct s390_idle_data, s390_idle);
+
 static void __init smp_create_idle(unsigned int cpu)
 {
 	struct task_struct *p;
@@ -506,6 +509,7 @@ static void __init smp_create_idle(unsigned int cpu)
 	if (IS_ERR(p))
 		panic("failed fork for CPU %u: %li", cpu, PTR_ERR(p));
 	current_set[cpu] = p;
+	spin_lock_init(&(&per_cpu(s390_idle, cpu))->lock);
 }
 
 static int cpu_stopped(int cpu)
@@ -724,6 +728,7 @@ void __init smp_prepare_boot_cpu(void)
 	cpu_set(0, cpu_online_map);
 	S390_lowcore.percpu_offset = __per_cpu_offset[0];
 	current_set[0] = current;
+	spin_lock_init(&(&__get_cpu_var(s390_idle))->lock);
 }
 
 void __init smp_cpus_done(unsigned int max_cpus)
@@ -756,22 +761,71 @@ static ssize_t show_capability(struct sys_device *dev, char *buf)
 }
 static SYSDEV_ATTR(capability, 0444, show_capability, NULL);
 
+static ssize_t show_idle_count(struct sys_device *dev, char *buf)
+{
+	struct s390_idle_data *idle;
+	unsigned long long idle_count;
+
+	idle = &per_cpu(s390_idle, dev->id);
+	spin_lock_irq(&idle->lock);
+	idle_count = idle->idle_count;
+	spin_unlock_irq(&idle->lock);
+	return sprintf(buf, "%llu\n", idle_count);
+}
+static SYSDEV_ATTR(idle_count, 0444, show_idle_count, NULL);
+
+static ssize_t show_idle_time(struct sys_device *dev, char *buf)
+{
+	struct s390_idle_data *idle;
+	unsigned long long new_time;
+
+	idle = &per_cpu(s390_idle, dev->id);
+	spin_lock_irq(&idle->lock);
+	if (idle->in_idle) {
+		new_time = get_clock();
+		idle->idle_time += new_time - idle->idle_enter;
+		idle->idle_enter = new_time;
+	}
+	new_time = idle->idle_time;
+	spin_unlock_irq(&idle->lock);
+	return sprintf(buf, "%llu us\n", new_time >> 12);
+}
+static SYSDEV_ATTR(idle_time, 0444, show_idle_time, NULL);
+
+static struct attribute *cpu_attrs[] = {
+	&attr_capability.attr,
+	&attr_idle_count.attr,
+	&attr_idle_time.attr,
+	NULL,
+};
+
+static struct attribute_group cpu_attr_group = {
+	.attrs = cpu_attrs,
+};
+
 static int __cpuinit smp_cpu_notify(struct notifier_block *self,
 				    unsigned long action, void *hcpu)
 {
 	unsigned int cpu = (unsigned int)(long)hcpu;
 	struct cpu *c = &per_cpu(cpu_devices, cpu);
 	struct sys_device *s = &c->sysdev;
+	struct s390_idle_data *idle;
 
 	switch (action) {
 	case CPU_ONLINE:
 	case CPU_ONLINE_FROZEN:
-		if (sysdev_create_file(s, &attr_capability))
+		idle = &per_cpu(s390_idle, cpu);
+		spin_lock_irq(&idle->lock);
+		idle->idle_enter = 0;
+		idle->idle_time = 0;
+		idle->idle_count = 0;
+		spin_unlock_irq(&idle->lock);
+		if (sysfs_create_group(&s->kobj, &cpu_attr_group))
 			return NOTIFY_BAD;
 		break;
 	case CPU_DEAD:
 	case CPU_DEAD_FROZEN:
-		sysdev_remove_file(s, &attr_capability);
+		sysfs_remove_group(&s->kobj, &cpu_attr_group);
 		break;
 	}
 	return NOTIFY_OK;
@@ -784,6 +838,7 @@ static struct notifier_block __cpuinitdata smp_cpu_nb = {
 static int __init topology_init(void)
 {
 	int cpu;
+	int rc;
 
 	register_cpu_notifier(&smp_cpu_nb);
 
@@ -796,7 +851,9 @@ static int __init topology_init(void)
 		if (!cpu_online(cpu))
 			continue;
 		s = &c->sysdev;
-		sysdev_create_file(s, &attr_capability);
+		rc = sysfs_create_group(&s->kobj, &cpu_attr_group);
+		if (rc)
+			return rc;
 	}
 	return 0;
 }
diff --git a/arch/s390/lib/uaccess_pt.c b/arch/s390/lib/uaccess_pt.c
index b159a9d6568..7e8efaade2e 100644
--- a/arch/s390/lib/uaccess_pt.c
+++ b/arch/s390/lib/uaccess_pt.c
@@ -15,6 +15,27 @@
 #include <asm/futex.h>
 #include "uaccess.h"
 
+static inline pte_t *follow_table(struct mm_struct *mm, unsigned long addr)
+{
+	pgd_t *pgd;
+	pud_t *pud;
+	pmd_t *pmd;
+
+	pgd = pgd_offset(mm, addr);
+	if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd)))
+		return NULL;
+
+	pud = pud_offset(pgd, addr);
+	if (pud_none(*pud) || unlikely(pud_bad(*pud)))
+		return NULL;
+
+	pmd = pmd_offset(pud, addr);
+	if (pmd_none(*pmd) || unlikely(pmd_bad(*pmd)))
+		return NULL;
+
+	return pte_offset_map(pmd, addr);
+}
+
 static int __handle_fault(struct mm_struct *mm, unsigned long address,
 			  int write_access)
 {
@@ -85,8 +106,6 @@ static size_t __user_copy_pt(unsigned long uaddr, void *kptr,
 {
 	struct mm_struct *mm = current->mm;
 	unsigned long offset, pfn, done, size;
-	pgd_t *pgd;
-	pmd_t *pmd;
 	pte_t *pte;
 	void *from, *to;
 
@@ -94,15 +113,7 @@ static size_t __user_copy_pt(unsigned long uaddr, void *kptr,
 retry:
 	spin_lock(&mm->page_table_lock);
 	do {
-		pgd = pgd_offset(mm, uaddr);
-		if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd)))
-			goto fault;
-
-		pmd = pmd_offset(pgd, uaddr);
-		if (pmd_none(*pmd) || unlikely(pmd_bad(*pmd)))
-			goto fault;
-
-		pte = pte_offset_map(pmd, uaddr);
+		pte = follow_table(mm, uaddr);
 		if (!pte || !pte_present(*pte) ||
 		    (write_user && !pte_write(*pte)))
 			goto fault;
@@ -142,22 +153,12 @@ static unsigned long __dat_user_addr(unsigned long uaddr)
 {
 	struct mm_struct *mm = current->mm;
 	unsigned long pfn, ret;
-	pgd_t *pgd;
-	pmd_t *pmd;
 	pte_t *pte;
 	int rc;
 
 	ret = 0;
 retry:
-	pgd = pgd_offset(mm, uaddr);
-	if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd)))
-		goto fault;
-
-	pmd = pmd_offset(pgd, uaddr);
-	if (pmd_none(*pmd) || unlikely(pmd_bad(*pmd)))
-		goto fault;
-
-	pte = pte_offset_map(pmd, uaddr);
+	pte = follow_table(mm, uaddr);
 	if (!pte || !pte_present(*pte))
 		goto fault;
 
@@ -229,8 +230,6 @@ static size_t strnlen_user_pt(size_t count, const char __user *src)
 	unsigned long uaddr = (unsigned long) src;
 	struct mm_struct *mm = current->mm;
 	unsigned long offset, pfn, done, len;
-	pgd_t *pgd;
-	pmd_t *pmd;
 	pte_t *pte;
 	size_t len_str;
 
@@ -240,15 +239,7 @@ static size_t strnlen_user_pt(size_t count, const char __user *src)
 retry:
 	spin_lock(&mm->page_table_lock);
 	do {
-		pgd = pgd_offset(mm, uaddr);
-		if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd)))
-			goto fault;
-
-		pmd = pmd_offset(pgd, uaddr);
-		if (pmd_none(*pmd) || unlikely(pmd_bad(*pmd)))
-			goto fault;
-
-		pte = pte_offset_map(pmd, uaddr);
+		pte = follow_table(mm, uaddr);
 		if (!pte || !pte_present(*pte))
 			goto fault;
 
@@ -308,8 +299,6 @@ static size_t copy_in_user_pt(size_t n, void __user *to,
 		      uaddr, done, size;
 	unsigned long uaddr_from = (unsigned long) from;
 	unsigned long uaddr_to = (unsigned long) to;
-	pgd_t *pgd_from, *pgd_to;
-	pmd_t *pmd_from, *pmd_to;
 	pte_t *pte_from, *pte_to;
 	int write_user;
 
@@ -317,39 +306,14 @@ static size_t copy_in_user_pt(size_t n, void __user *to,
 retry:
 	spin_lock(&mm->page_table_lock);
 	do {
-		pgd_from = pgd_offset(mm, uaddr_from);
-		if (pgd_none(*pgd_from) || unlikely(pgd_bad(*pgd_from))) {
-			uaddr = uaddr_from;
-			write_user = 0;
-			goto fault;
-		}
-		pgd_to = pgd_offset(mm, uaddr_to);
-		if (pgd_none(*pgd_to) || unlikely(pgd_bad(*pgd_to))) {
-			uaddr = uaddr_to;
-			write_user = 1;
-			goto fault;
-		}
-
-		pmd_from = pmd_offset(pgd_from, uaddr_from);
-		if (pmd_none(*pmd_from) || unlikely(pmd_bad(*pmd_from))) {
-			uaddr = uaddr_from;
-			write_user = 0;
-			goto fault;
-		}
-		pmd_to = pmd_offset(pgd_to, uaddr_to);
-		if (pmd_none(*pmd_to) || unlikely(pmd_bad(*pmd_to))) {
-			uaddr = uaddr_to;
-			write_user = 1;
-			goto fault;
-		}
-
-		pte_from = pte_offset_map(pmd_from, uaddr_from);
+		pte_from = follow_table(mm, uaddr_from);
 		if (!pte_from || !pte_present(*pte_from)) {
 			uaddr = uaddr_from;
 			write_user = 0;
 			goto fault;
 		}
-		pte_to = pte_offset_map(pmd_to, uaddr_to);
+
+		pte_to = follow_table(mm, uaddr_to);
 		if (!pte_to || !pte_present(*pte_to) || !pte_write(*pte_to)) {
 			uaddr = uaddr_to;
 			write_user = 1;
diff --git a/arch/s390/mm/Makefile b/arch/s390/mm/Makefile
index f95449b29fa..66401930f83 100644
--- a/arch/s390/mm/Makefile
+++ b/arch/s390/mm/Makefile
@@ -2,6 +2,6 @@
 # Makefile for the linux s390-specific parts of the memory manager.
 #
 
-obj-y	 := init.o fault.o extmem.o mmap.o vmem.o
+obj-y	 := init.o fault.o extmem.o mmap.o vmem.o pgtable.o
 obj-$(CONFIG_CMM) += cmm.o
 
diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c
index 3a25bbf2eb0..b234bb4a6da 100644
--- a/arch/s390/mm/init.c
+++ b/arch/s390/mm/init.c
@@ -81,6 +81,7 @@ void show_mem(void)
 static void __init setup_ro_region(void)
 {
 	pgd_t *pgd;
+	pud_t *pud;
 	pmd_t *pmd;
 	pte_t *pte;
 	pte_t new_pte;
@@ -91,7 +92,8 @@ static void __init setup_ro_region(void)
 
 	for (; address < end; address += PAGE_SIZE) {
 		pgd = pgd_offset_k(address);
-		pmd = pmd_offset(pgd, address);
+		pud = pud_offset(pgd, address);
+		pmd = pmd_offset(pud, address);
 		pte = pte_offset_kernel(pmd, address);
 		new_pte = mk_pte_phys(address, __pgprot(_PAGE_RO));
 		*pte = new_pte;
@@ -103,32 +105,28 @@ static void __init setup_ro_region(void)
  */
 void __init paging_init(void)
 {
-	pgd_t *pg_dir;
-	int i;
-	unsigned long pgdir_k;
 	static const int ssm_mask = 0x04000000L;
 	unsigned long max_zone_pfns[MAX_NR_ZONES];
+	unsigned long pgd_type;
 
-	pg_dir = swapper_pg_dir;
-	
+	init_mm.pgd = swapper_pg_dir;
+	S390_lowcore.kernel_asce = __pa(init_mm.pgd) & PAGE_MASK;
 #ifdef CONFIG_64BIT
-	pgdir_k = (__pa(swapper_pg_dir) & PAGE_MASK) | _KERN_REGION_TABLE;
-	for (i = 0; i < PTRS_PER_PGD; i++)
-		pgd_clear_kernel(pg_dir + i);
+	S390_lowcore.kernel_asce |= _ASCE_TYPE_REGION3 | _ASCE_TABLE_LENGTH;
+	pgd_type = _REGION3_ENTRY_EMPTY;
 #else
-	pgdir_k = (__pa(swapper_pg_dir) & PAGE_MASK) | _KERNSEG_TABLE;
-	for (i = 0; i < PTRS_PER_PGD; i++)
-		pmd_clear_kernel((pmd_t *)(pg_dir + i));
+	S390_lowcore.kernel_asce |= _ASCE_TABLE_LENGTH;
+	pgd_type = _SEGMENT_ENTRY_EMPTY;
 #endif
+	clear_table((unsigned long *) init_mm.pgd, pgd_type,
+		    sizeof(unsigned long)*2048);
 	vmem_map_init();
 	setup_ro_region();
 
-	S390_lowcore.kernel_asce = pgdir_k;
-
         /* enable virtual mapping in kernel mode */
-	__ctl_load(pgdir_k, 1, 1);
-	__ctl_load(pgdir_k, 7, 7);
-	__ctl_load(pgdir_k, 13, 13);
+	__ctl_load(S390_lowcore.kernel_asce, 1, 1);
+	__ctl_load(S390_lowcore.kernel_asce, 7, 7);
+	__ctl_load(S390_lowcore.kernel_asce, 13, 13);
 	__raw_local_irq_ssm(ssm_mask);
 
 	memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c
new file mode 100644
index 00000000000..e60e0ae1340
--- /dev/null
+++ b/arch/s390/mm/pgtable.c
@@ -0,0 +1,94 @@
+/*
+ *  arch/s390/mm/pgtable.c
+ *
+ *    Copyright IBM Corp. 2007
+ *    Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/mm.h>
+#include <linux/swap.h>
+#include <linux/smp.h>
+#include <linux/highmem.h>
+#include <linux/slab.h>
+#include <linux/pagemap.h>
+#include <linux/spinlock.h>
+#include <linux/module.h>
+#include <linux/quicklist.h>
+
+#include <asm/system.h>
+#include <asm/pgtable.h>
+#include <asm/pgalloc.h>
+#include <asm/tlb.h>
+#include <asm/tlbflush.h>
+
+#ifndef CONFIG_64BIT
+#define ALLOC_ORDER	1
+#else
+#define ALLOC_ORDER	2
+#endif
+
+unsigned long *crst_table_alloc(struct mm_struct *mm, int noexec)
+{
+	struct page *page = alloc_pages(GFP_KERNEL, ALLOC_ORDER);
+
+	if (!page)
+		return NULL;
+	page->index = 0;
+	if (noexec) {
+		struct page *shadow = alloc_pages(GFP_KERNEL, ALLOC_ORDER);
+		if (!shadow) {
+			__free_pages(page, ALLOC_ORDER);
+			return NULL;
+		}
+		page->index = page_to_phys(shadow);
+	}
+	return (unsigned long *) page_to_phys(page);
+}
+
+void crst_table_free(unsigned long *table)
+{
+	unsigned long *shadow = get_shadow_table(table);
+
+	if (shadow)
+		free_pages((unsigned long) shadow, ALLOC_ORDER);
+	free_pages((unsigned long) table, ALLOC_ORDER);
+}
+
+/*
+ * page table entry allocation/free routines.
+ */
+unsigned long *page_table_alloc(int noexec)
+{
+	struct page *page = alloc_page(GFP_KERNEL);
+	unsigned long *table;
+
+	if (!page)
+		return NULL;
+	page->index = 0;
+	if (noexec) {
+		struct page *shadow = alloc_page(GFP_KERNEL);
+		if (!shadow) {
+			__free_page(page);
+			return NULL;
+		}
+		table = (unsigned long *) page_to_phys(shadow);
+		clear_table(table, _PAGE_TYPE_EMPTY, PAGE_SIZE);
+		page->index = (addr_t) table;
+	}
+	table = (unsigned long *) page_to_phys(page);
+	clear_table(table, _PAGE_TYPE_EMPTY, PAGE_SIZE);
+	return table;
+}
+
+void page_table_free(unsigned long *table)
+{
+	unsigned long *shadow = get_shadow_pte(table);
+
+	if (shadow)
+		free_page((unsigned long) shadow);
+	free_page((unsigned long) table);
+
+}
diff --git a/arch/s390/mm/vmem.c b/arch/s390/mm/vmem.c
index fd594d5fe14..fb9c5a85aa5 100644
--- a/arch/s390/mm/vmem.c
+++ b/arch/s390/mm/vmem.c
@@ -73,31 +73,28 @@ static void __init_refok *vmem_alloc_pages(unsigned int order)
 	return alloc_bootmem_pages((1 << order) * PAGE_SIZE);
 }
 
+#define vmem_pud_alloc()	({ BUG(); ((pud_t *) NULL); })
+
 static inline pmd_t *vmem_pmd_alloc(void)
 {
-	pmd_t *pmd;
-	int i;
+	pmd_t *pmd = NULL;
 
-	pmd = vmem_alloc_pages(PMD_ALLOC_ORDER);
+#ifdef CONFIG_64BIT
+	pmd = vmem_alloc_pages(2);
 	if (!pmd)
 		return NULL;
-	for (i = 0; i < PTRS_PER_PMD; i++)
-		pmd_clear_kernel(pmd + i);
+	clear_table((unsigned long *) pmd, _SEGMENT_ENTRY_EMPTY, PAGE_SIZE*4);
+#endif
 	return pmd;
 }
 
 static inline pte_t *vmem_pte_alloc(void)
 {
-	pte_t *pte;
-	pte_t empty_pte;
-	int i;
+	pte_t *pte = vmem_alloc_pages(0);
 
-	pte = vmem_alloc_pages(PTE_ALLOC_ORDER);
 	if (!pte)
 		return NULL;
-	pte_val(empty_pte) = _PAGE_TYPE_EMPTY;
-	for (i = 0; i < PTRS_PER_PTE; i++)
-		pte[i] = empty_pte;
+	clear_table((unsigned long *) pte, _PAGE_TYPE_EMPTY, PAGE_SIZE);
 	return pte;
 }
 
@@ -108,6 +105,7 @@ static int vmem_add_range(unsigned long start, unsigned long size)
 {
 	unsigned long address;
 	pgd_t *pg_dir;
+	pud_t *pu_dir;
 	pmd_t *pm_dir;
 	pte_t *pt_dir;
 	pte_t  pte;
@@ -116,13 +114,21 @@ static int vmem_add_range(unsigned long start, unsigned long size)
 	for (address = start; address < start + size; address += PAGE_SIZE) {
 		pg_dir = pgd_offset_k(address);
 		if (pgd_none(*pg_dir)) {
+			pu_dir = vmem_pud_alloc();
+			if (!pu_dir)
+				goto out;
+			pgd_populate_kernel(&init_mm, pg_dir, pu_dir);
+		}
+
+		pu_dir = pud_offset(pg_dir, address);
+		if (pud_none(*pu_dir)) {
 			pm_dir = vmem_pmd_alloc();
 			if (!pm_dir)
 				goto out;
-			pgd_populate_kernel(&init_mm, pg_dir, pm_dir);
+			pud_populate_kernel(&init_mm, pu_dir, pm_dir);
 		}
 
-		pm_dir = pmd_offset(pg_dir, address);
+		pm_dir = pmd_offset(pu_dir, address);
 		if (pmd_none(*pm_dir)) {
 			pt_dir = vmem_pte_alloc();
 			if (!pt_dir)
@@ -148,6 +154,7 @@ static void vmem_remove_range(unsigned long start, unsigned long size)
 {
 	unsigned long address;
 	pgd_t *pg_dir;
+	pud_t *pu_dir;
 	pmd_t *pm_dir;
 	pte_t *pt_dir;
 	pte_t  pte;
@@ -155,9 +162,10 @@ static void vmem_remove_range(unsigned long start, unsigned long size)
 	pte_val(pte) = _PAGE_TYPE_EMPTY;
 	for (address = start; address < start + size; address += PAGE_SIZE) {
 		pg_dir = pgd_offset_k(address);
-		if (pgd_none(*pg_dir))
+		pu_dir = pud_offset(pg_dir, address);
+		if (pud_none(*pu_dir))
 			continue;
-		pm_dir = pmd_offset(pg_dir, address);
+		pm_dir = pmd_offset(pu_dir, address);
 		if (pmd_none(*pm_dir))
 			continue;
 		pt_dir = pte_offset_kernel(pm_dir, address);
@@ -174,6 +182,7 @@ static int vmem_add_mem_map(unsigned long start, unsigned long size)
 	unsigned long address, start_addr, end_addr;
 	struct page *map_start, *map_end;
 	pgd_t *pg_dir;
+	pud_t *pu_dir;
 	pmd_t *pm_dir;
 	pte_t *pt_dir;
 	pte_t  pte;
@@ -188,13 +197,21 @@ static int vmem_add_mem_map(unsigned long start, unsigned long size)
 	for (address = start_addr; address < end_addr; address += PAGE_SIZE) {
 		pg_dir = pgd_offset_k(address);
 		if (pgd_none(*pg_dir)) {
+			pu_dir = vmem_pud_alloc();
+			if (!pu_dir)
+				goto out;
+			pgd_populate_kernel(&init_mm, pg_dir, pu_dir);
+		}
+
+		pu_dir = pud_offset(pg_dir, address);
+		if (pud_none(*pu_dir)) {
 			pm_dir = vmem_pmd_alloc();
 			if (!pm_dir)
 				goto out;
-			pgd_populate_kernel(&init_mm, pg_dir, pm_dir);
+			pud_populate_kernel(&init_mm, pu_dir, pm_dir);
 		}
 
-		pm_dir = pmd_offset(pg_dir, address);
+		pm_dir = pmd_offset(pu_dir, address);
 		if (pmd_none(*pm_dir)) {
 			pt_dir = vmem_pte_alloc();
 			if (!pt_dir)
diff --git a/drivers/s390/char/raw3270.c b/drivers/s390/char/raw3270.c
index 2edd5fb6d3d..8d1c64a24de 100644
--- a/drivers/s390/char/raw3270.c
+++ b/drivers/s390/char/raw3270.c
@@ -48,8 +48,8 @@ struct raw3270 {
 	struct timer_list timer;	/* Device timer. */
 
 	unsigned char *ascebc;		/* ascii -> ebcdic table */
-	struct class_device *clttydev;	/* 3270-class tty device ptr */
-	struct class_device *cltubdev;	/* 3270-class tub device ptr */
+	struct device *clttydev;	/* 3270-class tty device ptr */
+	struct device *cltubdev;	/* 3270-class tub device ptr */
 
 	struct raw3270_request init_request;
 	unsigned char init_data[256];
@@ -1107,11 +1107,9 @@ raw3270_delete_device(struct raw3270 *rp)
 	/* Remove from device chain. */
 	mutex_lock(&raw3270_mutex);
 	if (rp->clttydev && !IS_ERR(rp->clttydev))
-		class_device_destroy(class3270,
-				     MKDEV(IBM_TTY3270_MAJOR, rp->minor));
+		device_destroy(class3270, MKDEV(IBM_TTY3270_MAJOR, rp->minor));
 	if (rp->cltubdev && !IS_ERR(rp->cltubdev))
-		class_device_destroy(class3270,
-				     MKDEV(IBM_FS3270_MAJOR, rp->minor));
+		device_destroy(class3270, MKDEV(IBM_FS3270_MAJOR, rp->minor));
 	list_del_init(&rp->list);
 	mutex_unlock(&raw3270_mutex);
 
@@ -1181,24 +1179,22 @@ static int raw3270_create_attributes(struct raw3270 *rp)
 	if (rc)
 		goto out;
 
-	rp->clttydev = class_device_create(class3270, NULL,
-					   MKDEV(IBM_TTY3270_MAJOR, rp->minor),
-					   &rp->cdev->dev, "tty%s",
-					   rp->cdev->dev.bus_id);
+	rp->clttydev = device_create(class3270, &rp->cdev->dev,
+				     MKDEV(IBM_TTY3270_MAJOR, rp->minor),
+				     "tty%s", rp->cdev->dev.bus_id);
 	if (IS_ERR(rp->clttydev)) {
 		rc = PTR_ERR(rp->clttydev);
 		goto out_ttydev;
 	}
 
-	rp->cltubdev = class_device_create(class3270, NULL,
-					   MKDEV(IBM_FS3270_MAJOR, rp->minor),
-					   &rp->cdev->dev, "tub%s",
-					   rp->cdev->dev.bus_id);
+	rp->cltubdev = device_create(class3270, &rp->cdev->dev,
+				     MKDEV(IBM_FS3270_MAJOR, rp->minor),
+				     "tub%s", rp->cdev->dev.bus_id);
 	if (!IS_ERR(rp->cltubdev))
 		goto out;
 
 	rc = PTR_ERR(rp->cltubdev);
-	class_device_destroy(class3270, MKDEV(IBM_TTY3270_MAJOR, rp->minor));
+	device_destroy(class3270, MKDEV(IBM_TTY3270_MAJOR, rp->minor));
 
 out_ttydev:
 	sysfs_remove_group(&rp->cdev->dev.kobj, &raw3270_attr_group);
diff --git a/drivers/s390/char/tape_class.c b/drivers/s390/char/tape_class.c
index 2e0d29730b6..aa7f166f403 100644
--- a/drivers/s390/char/tape_class.c
+++ b/drivers/s390/char/tape_class.c
@@ -69,12 +69,9 @@ struct tape_class_device *register_tape_dev(
 	if (rc)
 		goto fail_with_cdev;
 
-	tcd->class_device = class_device_create(
-				tape_class,
-				NULL,
-				tcd->char_device->dev,
-				device,
-				"%s", tcd->device_name
+	tcd->class_device = device_create(tape_class, device,
+					  tcd->char_device->dev,
+					  "%s", tcd->device_name
 			);
 	rc = IS_ERR(tcd->class_device) ? PTR_ERR(tcd->class_device) : 0;
 	if (rc)
@@ -90,7 +87,7 @@ struct tape_class_device *register_tape_dev(
 	return tcd;
 
 fail_with_class_device:
-	class_device_destroy(tape_class, tcd->char_device->dev);
+	device_destroy(tape_class, tcd->char_device->dev);
 
 fail_with_cdev:
 	cdev_del(tcd->char_device);
@@ -105,11 +102,9 @@ EXPORT_SYMBOL(register_tape_dev);
 void unregister_tape_dev(struct tape_class_device *tcd)
 {
 	if (tcd != NULL && !IS_ERR(tcd)) {
-		sysfs_remove_link(
-			&tcd->class_device->dev->kobj,
-			tcd->mode_name
-		);
-		class_device_destroy(tape_class, tcd->char_device->dev);
+		sysfs_remove_link(&tcd->class_device->kobj,
+				  tcd->mode_name);
+		device_destroy(tape_class, tcd->char_device->dev);
 		cdev_del(tcd->char_device);
 		kfree(tcd);
 	}
diff --git a/drivers/s390/char/tape_class.h b/drivers/s390/char/tape_class.h
index a8bd9b47fad..e2b5ac918ac 100644
--- a/drivers/s390/char/tape_class.h
+++ b/drivers/s390/char/tape_class.h
@@ -24,8 +24,8 @@
 #define TAPECLASS_NAME_LEN	32
 
 struct tape_class_device {
-	struct cdev *		char_device;
-	struct class_device *	class_device;
+	struct cdev		*char_device;
+	struct device		*class_device;
 	char			device_name[TAPECLASS_NAME_LEN];
 	char			mode_name[TAPECLASS_NAME_LEN];
 };
diff --git a/drivers/s390/char/vmlogrdr.c b/drivers/s390/char/vmlogrdr.c
index 12f7a4ce82c..e0c4c508e12 100644
--- a/drivers/s390/char/vmlogrdr.c
+++ b/drivers/s390/char/vmlogrdr.c
@@ -74,7 +74,7 @@ struct vmlogrdr_priv_t {
 	int dev_in_use; /* 1: already opened, 0: not opened*/
 	spinlock_t priv_lock;
 	struct device  *device;
-	struct class_device  *class_device;
+	struct device  *class_device;
 	int autorecording;
 	int autopurge;
 };
@@ -762,12 +762,10 @@ static int vmlogrdr_register_device(struct vmlogrdr_priv_t *priv)
 		device_unregister(dev);
 		return ret;
 	}
-	priv->class_device = class_device_create(
-				vmlogrdr_class,
-				NULL,
-				MKDEV(vmlogrdr_major, priv->minor_num),
-				dev,
-				"%s", dev->bus_id );
+	priv->class_device = device_create(vmlogrdr_class, dev,
+					   MKDEV(vmlogrdr_major,
+						 priv->minor_num),
+					   "%s", dev->bus_id);
 	if (IS_ERR(priv->class_device)) {
 		ret = PTR_ERR(priv->class_device);
 		priv->class_device=NULL;
@@ -783,8 +781,7 @@ static int vmlogrdr_register_device(struct vmlogrdr_priv_t *priv)
 
 static int vmlogrdr_unregister_device(struct vmlogrdr_priv_t *priv)
 {
-	class_device_destroy(vmlogrdr_class,
-			     MKDEV(vmlogrdr_major, priv->minor_num));
+	device_destroy(vmlogrdr_class, MKDEV(vmlogrdr_major, priv->minor_num));
 	if (priv->device != NULL) {
 		sysfs_remove_group(&priv->device->kobj, &vmlogrdr_attr_group);
 		device_unregister(priv->device);
diff --git a/drivers/s390/cio/chp.c b/drivers/s390/cio/chp.c
index 42c1f4659ad..297cdceb0ca 100644
--- a/drivers/s390/cio/chp.c
+++ b/drivers/s390/cio/chp.c
@@ -246,7 +246,7 @@ int chp_add_cmg_attr(struct channel_path *chp)
 static ssize_t chp_status_show(struct device *dev,
 			       struct device_attribute *attr, char *buf)
 {
-	struct channel_path *chp = container_of(dev, struct channel_path, dev);
+	struct channel_path *chp = to_channelpath(dev);
 
 	if (!chp)
 		return 0;
@@ -258,7 +258,7 @@ static ssize_t chp_status_write(struct device *dev,
 				struct device_attribute *attr,
 				const char *buf, size_t count)
 {
-	struct channel_path *cp = container_of(dev, struct channel_path, dev);
+	struct channel_path *cp = to_channelpath(dev);
 	char cmd[10];
 	int num_args;
 	int error;
@@ -286,7 +286,7 @@ static ssize_t chp_configure_show(struct device *dev,
 	struct channel_path *cp;
 	int status;
 
-	cp = container_of(dev, struct channel_path, dev);
+	cp = to_channelpath(dev);
 	status = chp_info_get_status(cp->chpid);
 	if (status < 0)
 		return status;
@@ -308,7 +308,7 @@ static ssize_t chp_configure_write(struct device *dev,
 		return -EINVAL;
 	if (val != 0 && val != 1)
 		return -EINVAL;
-	cp = container_of(dev, struct channel_path, dev);
+	cp = to_channelpath(dev);
 	chp_cfg_schedule(cp->chpid, val);
 	cfg_wait_idle();
 
@@ -320,7 +320,7 @@ static DEVICE_ATTR(configure, 0644, chp_configure_show, chp_configure_write);
 static ssize_t chp_type_show(struct device *dev, struct device_attribute *attr,
 			     char *buf)
 {
-	struct channel_path *chp = container_of(dev, struct channel_path, dev);
+	struct channel_path *chp = to_channelpath(dev);
 
 	if (!chp)
 		return 0;
@@ -374,7 +374,7 @@ static void chp_release(struct device *dev)
 {
 	struct channel_path *cp;
 
-	cp = container_of(dev, struct channel_path, dev);
+	cp = to_channelpath(dev);
 	kfree(cp);
 }
 
diff --git a/drivers/s390/cio/css.c b/drivers/s390/cio/css.c
index 5d83dd47146..838f7ac0dc3 100644
--- a/drivers/s390/cio/css.c
+++ b/drivers/s390/cio/css.c
@@ -182,6 +182,15 @@ static int css_register_subchannel(struct subchannel *sch)
 	sch->dev.bus = &css_bus_type;
 	sch->dev.release = &css_subchannel_release;
 	sch->dev.groups = subch_attr_groups;
+	/*
+	 * We don't want to generate uevents for I/O subchannels that don't
+	 * have a working ccw device behind them since they will be
+	 * unregistered before they can be used anyway, so we delay the add
+	 * uevent until after device recognition was successful.
+	 */
+	if (!cio_is_console(sch->schid))
+		/* Console is special, no need to suppress. */
+		sch->dev.uevent_suppress = 1;
 	css_update_ssd_info(sch);
 	/* make it known to the system */
 	ret = css_sch_device_register(sch);
diff --git a/include/asm-s390/cpu.h b/include/asm-s390/cpu.h
new file mode 100644
index 00000000000..352dde194f3
--- /dev/null
+++ b/include/asm-s390/cpu.h
@@ -0,0 +1,25 @@
+/*
+ *  include/asm-s390/cpu.h
+ *
+ *    Copyright IBM Corp. 2007
+ *    Author(s): Heiko Carstens <heiko.carstens@de.ibm.com>
+ */
+
+#ifndef _ASM_S390_CPU_H_
+#define _ASM_S390_CPU_H_
+
+#include <linux/types.h>
+#include <linux/percpu.h>
+#include <linux/spinlock.h>
+
+struct s390_idle_data {
+	spinlock_t lock;
+	unsigned int in_idle;
+	unsigned long long idle_count;
+	unsigned long long idle_enter;
+	unsigned long long idle_time;
+};
+
+DECLARE_PER_CPU(struct s390_idle_data, s390_idle);
+
+#endif /* _ASM_S390_CPU_H_ */
diff --git a/include/asm-s390/mmu_context.h b/include/asm-s390/mmu_context.h
index 501cb9b0631..05b842126b9 100644
--- a/include/asm-s390/mmu_context.h
+++ b/include/asm-s390/mmu_context.h
@@ -21,45 +21,43 @@
 
 #ifndef __s390x__
 #define LCTL_OPCODE "lctl"
-#define PGTABLE_BITS (_SEGMENT_TABLE|USER_STD_MASK)
 #else
 #define LCTL_OPCODE "lctlg"
-#define PGTABLE_BITS (_REGION_TABLE|USER_STD_MASK)
 #endif
 
-static inline void enter_lazy_tlb(struct mm_struct *mm,
-                                  struct task_struct *tsk)
+static inline void update_mm(struct mm_struct *mm, struct task_struct *tsk)
 {
+	pgd_t *pgd = mm->pgd;
+	unsigned long asce_bits;
+
+	/* Calculate asce bits from the first pgd table entry. */
+	asce_bits = _ASCE_TABLE_LENGTH | _ASCE_USER_BITS;
+#ifdef CONFIG_64BIT
+	asce_bits |= _ASCE_TYPE_REGION3;
+#endif
+	S390_lowcore.user_asce = asce_bits | __pa(pgd);
+	if (switch_amode) {
+		/* Load primary space page table origin. */
+		pgd_t *shadow_pgd = get_shadow_table(pgd) ? : pgd;
+		S390_lowcore.user_exec_asce = asce_bits | __pa(shadow_pgd);
+		asm volatile(LCTL_OPCODE" 1,1,%0\n"
+			     : : "m" (S390_lowcore.user_exec_asce) );
+	} else
+		/* Load home space page table origin. */
+		asm volatile(LCTL_OPCODE" 13,13,%0"
+			     : : "m" (S390_lowcore.user_asce) );
 }
 
 static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next,
 			     struct task_struct *tsk)
 {
-	pgd_t *shadow_pgd = get_shadow_pgd(next->pgd);
-
-	if (prev != next) {
-		S390_lowcore.user_asce = (__pa(next->pgd) & PAGE_MASK) |
-					 PGTABLE_BITS;
-		if (shadow_pgd) {
-			/* Load primary/secondary space page table origin. */
-			S390_lowcore.user_exec_asce =
-				(__pa(shadow_pgd) & PAGE_MASK) | PGTABLE_BITS;
-			asm volatile(LCTL_OPCODE" 1,1,%0\n"
-				     LCTL_OPCODE" 7,7,%1"
-				     : : "m" (S390_lowcore.user_exec_asce),
-					 "m" (S390_lowcore.user_asce) );
-		} else if (switch_amode) {
-			/* Load primary space page table origin. */
-			asm volatile(LCTL_OPCODE" 1,1,%0"
-				     : : "m" (S390_lowcore.user_asce) );
-		} else
-			/* Load home space page table origin. */
-			asm volatile(LCTL_OPCODE" 13,13,%0"
-				     : : "m" (S390_lowcore.user_asce) );
-	}
+	if (unlikely(prev == next))
+		return;
 	cpu_set(smp_processor_id(), next->cpu_vm_mask);
+	update_mm(next, tsk);
 }
 
+#define enter_lazy_tlb(mm,tsk)	do { } while (0)
 #define deactivate_mm(tsk,mm)	do { } while (0)
 
 static inline void activate_mm(struct mm_struct *prev,
diff --git a/include/asm-s390/page.h b/include/asm-s390/page.h
index ceec3826a67..584d0ee3c7f 100644
--- a/include/asm-s390/page.h
+++ b/include/asm-s390/page.h
@@ -82,6 +82,7 @@ typedef struct { unsigned long pte; } pte_t;
 #ifndef __s390x__
 
 typedef struct { unsigned long pmd; } pmd_t;
+typedef struct { unsigned long pud; } pud_t;
 typedef struct {
         unsigned long pgd0;
         unsigned long pgd1;
@@ -90,6 +91,7 @@ typedef struct {
         } pgd_t;
 
 #define pmd_val(x)      ((x).pmd)
+#define pud_val(x)	((x).pud)
 #define pgd_val(x)      ((x).pgd0)
 
 #else /* __s390x__ */
@@ -98,10 +100,12 @@ typedef struct {
         unsigned long pmd0;
         unsigned long pmd1; 
         } pmd_t;
+typedef struct { unsigned long pud; } pud_t;
 typedef struct { unsigned long pgd; } pgd_t;
 
 #define pmd_val(x)      ((x).pmd0)
 #define pmd_val1(x)     ((x).pmd1)
+#define pud_val(x)	((x).pud)
 #define pgd_val(x)      ((x).pgd)
 
 #endif /* __s390x__ */
diff --git a/include/asm-s390/pgalloc.h b/include/asm-s390/pgalloc.h
index e45d3c9a4b7..709dd174095 100644
--- a/include/asm-s390/pgalloc.h
+++ b/include/asm-s390/pgalloc.h
@@ -19,140 +19,115 @@
 
 #define check_pgt_cache()	do {} while (0)
 
-/*
- * Page allocation orders.
- */
-#ifndef __s390x__
-# define PTE_ALLOC_ORDER	0
-# define PMD_ALLOC_ORDER	0
-# define PGD_ALLOC_ORDER	1
-#else /* __s390x__ */
-# define PTE_ALLOC_ORDER	0
-# define PMD_ALLOC_ORDER	2
-# define PGD_ALLOC_ORDER	2
-#endif /* __s390x__ */
+unsigned long *crst_table_alloc(struct mm_struct *, int);
+void crst_table_free(unsigned long *);
 
-/*
- * Allocate and free page tables. The xxx_kernel() versions are
- * used to allocate a kernel page table - this turns on ASN bits
- * if any.
- */
+unsigned long *page_table_alloc(int);
+void page_table_free(unsigned long *);
 
-static inline pgd_t *pgd_alloc(struct mm_struct *mm)
+static inline void clear_table(unsigned long *s, unsigned long val, size_t n)
 {
-	pgd_t *pgd = (pgd_t *) __get_free_pages(GFP_KERNEL, PGD_ALLOC_ORDER);
-	int i;
-
-	if (!pgd)
-		return NULL;
-	if (s390_noexec) {
-		pgd_t *shadow_pgd = (pgd_t *)
-			__get_free_pages(GFP_KERNEL, PGD_ALLOC_ORDER);
-		struct page *page = virt_to_page(pgd);
-
-		if (!shadow_pgd) {
-			free_pages((unsigned long) pgd, PGD_ALLOC_ORDER);
-			return NULL;
-		}
-		page->lru.next = (void *) shadow_pgd;
-	}
-	for (i = 0; i < PTRS_PER_PGD; i++)
-#ifndef __s390x__
-		pmd_clear(pmd_offset(pgd + i, i*PGDIR_SIZE));
+	*s = val;
+	n = (n / 256) - 1;
+	asm volatile(
+#ifdef CONFIG_64BIT
+		"	mvc	8(248,%0),0(%0)\n"
 #else
-		pgd_clear(pgd + i);
+		"	mvc	4(252,%0),0(%0)\n"
 #endif
-	return pgd;
+		"0:	mvc	256(256,%0),0(%0)\n"
+		"	la	%0,256(%0)\n"
+		"	brct	%1,0b\n"
+		: "+a" (s), "+d" (n));
 }
 
-static inline void pgd_free(pgd_t *pgd)
+static inline void crst_table_init(unsigned long *crst, unsigned long entry)
 {
-	pgd_t *shadow_pgd = get_shadow_pgd(pgd);
-
-	if (shadow_pgd)
-		free_pages((unsigned long) shadow_pgd, PGD_ALLOC_ORDER);
-	free_pages((unsigned long) pgd, PGD_ALLOC_ORDER);
+	clear_table(crst, entry, sizeof(unsigned long)*2048);
+	crst = get_shadow_table(crst);
+	if (crst)
+		clear_table(crst, entry, sizeof(unsigned long)*2048);
 }
 
 #ifndef __s390x__
-/*
- * page middle directory allocation/free routines.
- * We use pmd cache only on s390x, so these are dummy routines. This
- * code never triggers because the pgd will always be present.
- */
-#define pmd_alloc_one(mm,address)       ({ BUG(); ((pmd_t *)2); })
-#define pmd_free(x)                     do { } while (0)
-#define __pmd_free_tlb(tlb,x)		do { } while (0)
-#define pgd_populate(mm, pmd, pte)      BUG()
-#define pgd_populate_kernel(mm, pmd, pte)	BUG()
-#else /* __s390x__ */
-static inline pmd_t * pmd_alloc_one(struct mm_struct *mm, unsigned long vmaddr)
+
+static inline unsigned long pgd_entry_type(struct mm_struct *mm)
 {
-	pmd_t *pmd = (pmd_t *) __get_free_pages(GFP_KERNEL, PMD_ALLOC_ORDER);
-	int i;
-
-	if (!pmd)
-		return NULL;
-	if (s390_noexec) {
-		pmd_t *shadow_pmd = (pmd_t *)
-			__get_free_pages(GFP_KERNEL, PMD_ALLOC_ORDER);
-		struct page *page = virt_to_page(pmd);
-
-		if (!shadow_pmd) {
-			free_pages((unsigned long) pmd, PMD_ALLOC_ORDER);
-			return NULL;
-		}
-		page->lru.next = (void *) shadow_pmd;
-	}
-	for (i=0; i < PTRS_PER_PMD; i++)
-		pmd_clear(pmd + i);
-	return pmd;
+	return _SEGMENT_ENTRY_EMPTY;
 }
 
-static inline void pmd_free (pmd_t *pmd)
+#define pud_alloc_one(mm,address)		({ BUG(); ((pud_t *)2); })
+#define pud_free(x)				do { } while (0)
+
+#define pmd_alloc_one(mm,address)		({ BUG(); ((pmd_t *)2); })
+#define pmd_free(x)				do { } while (0)
+
+#define pgd_populate(mm, pgd, pud)		BUG()
+#define pgd_populate_kernel(mm, pgd, pud)	BUG()
+
+#define pud_populate(mm, pud, pmd)		BUG()
+#define pud_populate_kernel(mm, pud, pmd)	BUG()
+
+#else /* __s390x__ */
+
+static inline unsigned long pgd_entry_type(struct mm_struct *mm)
 {
-	pmd_t *shadow_pmd = get_shadow_pmd(pmd);
+	return _REGION3_ENTRY_EMPTY;
+}
+
+#define pud_alloc_one(mm,address)		({ BUG(); ((pud_t *)2); })
+#define pud_free(x)				do { } while (0)
 
-	if (shadow_pmd)
-		free_pages((unsigned long) shadow_pmd, PMD_ALLOC_ORDER);
-	free_pages((unsigned long) pmd, PMD_ALLOC_ORDER);
+static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long vmaddr)
+{
+	unsigned long *crst = crst_table_alloc(mm, s390_noexec);
+	if (crst)
+		crst_table_init(crst, _SEGMENT_ENTRY_EMPTY);
+	return (pmd_t *) crst;
 }
+#define pmd_free(pmd) crst_table_free((unsigned long *) pmd)
 
-#define __pmd_free_tlb(tlb,pmd)			\
-	do {					\
-		tlb_flush_mmu(tlb, 0, 0);	\
-		pmd_free(pmd);			\
-	 } while (0)
+#define pgd_populate(mm, pgd, pud)		BUG()
+#define pgd_populate_kernel(mm, pgd, pud)	BUG()
 
-static inline void
-pgd_populate_kernel(struct mm_struct *mm, pgd_t *pgd, pmd_t *pmd)
+static inline void pud_populate_kernel(struct mm_struct *mm,
+				       pud_t *pud, pmd_t *pmd)
 {
-	pgd_val(*pgd) = _PGD_ENTRY | __pa(pmd);
+	pud_val(*pud) = _REGION3_ENTRY | __pa(pmd);
 }
 
-static inline void pgd_populate(struct mm_struct *mm, pgd_t *pgd, pmd_t *pmd)
+static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd)
 {
-	pgd_t *shadow_pgd = get_shadow_pgd(pgd);
-	pmd_t *shadow_pmd = get_shadow_pmd(pmd);
+	pud_t *shadow_pud = get_shadow_table(pud);
+	pmd_t *shadow_pmd = get_shadow_table(pmd);
 
-	if (shadow_pgd && shadow_pmd)
-		pgd_populate_kernel(mm, shadow_pgd, shadow_pmd);
-	pgd_populate_kernel(mm, pgd, pmd);
+	if (shadow_pud && shadow_pmd)
+		pud_populate_kernel(mm, shadow_pud, shadow_pmd);
+	pud_populate_kernel(mm, pud, pmd);
 }
 
 #endif /* __s390x__ */
 
+static inline pgd_t *pgd_alloc(struct mm_struct *mm)
+{
+	unsigned long *crst = crst_table_alloc(mm, s390_noexec);
+	if (crst)
+		crst_table_init(crst, pgd_entry_type(mm));
+	return (pgd_t *) crst;
+}
+#define pgd_free(pgd) crst_table_free((unsigned long *) pgd)
+
 static inline void 
 pmd_populate_kernel(struct mm_struct *mm, pmd_t *pmd, pte_t *pte)
 {
 #ifndef __s390x__
-	pmd_val(pmd[0]) = _PAGE_TABLE + __pa(pte);
-	pmd_val(pmd[1]) = _PAGE_TABLE + __pa(pte+256);
-	pmd_val(pmd[2]) = _PAGE_TABLE + __pa(pte+512);
-	pmd_val(pmd[3]) = _PAGE_TABLE + __pa(pte+768);
+	pmd_val(pmd[0]) = _SEGMENT_ENTRY + __pa(pte);
+	pmd_val(pmd[1]) = _SEGMENT_ENTRY + __pa(pte+256);
+	pmd_val(pmd[2]) = _SEGMENT_ENTRY + __pa(pte+512);
+	pmd_val(pmd[3]) = _SEGMENT_ENTRY + __pa(pte+768);
 #else /* __s390x__ */
-	pmd_val(*pmd) = _PMD_ENTRY + __pa(pte);
-	pmd_val1(*pmd) = _PMD_ENTRY + __pa(pte+256);
+	pmd_val(*pmd) = _SEGMENT_ENTRY + __pa(pte);
+	pmd_val1(*pmd) = _SEGMENT_ENTRY + __pa(pte+256);
 #endif /* __s390x__ */
 }
 
@@ -160,7 +135,7 @@ static inline void
 pmd_populate(struct mm_struct *mm, pmd_t *pmd, struct page *page)
 {
 	pte_t *pte = (pte_t *)page_to_phys(page);
-	pmd_t *shadow_pmd = get_shadow_pmd(pmd);
+	pmd_t *shadow_pmd = get_shadow_table(pmd);
 	pte_t *shadow_pte = get_shadow_pte(pte);
 
 	pmd_populate_kernel(mm, pmd, pte);
@@ -171,67 +146,14 @@ pmd_populate(struct mm_struct *mm, pmd_t *pmd, struct page *page)
 /*
  * page table entry allocation/free routines.
  */
-static inline pte_t *
-pte_alloc_one_kernel(struct mm_struct *mm, unsigned long vmaddr)
-{
-	pte_t *pte = (pte_t *) __get_free_page(GFP_KERNEL|__GFP_REPEAT);
-	int i;
-
-	if (!pte)
-		return NULL;
-	if (s390_noexec) {
-		pte_t *shadow_pte = (pte_t *)
-			__get_free_page(GFP_KERNEL|__GFP_REPEAT);
-		struct page *page = virt_to_page(pte);
-
-		if (!shadow_pte) {
-			free_page((unsigned long) pte);
-			return NULL;
-		}
-		page->lru.next = (void *) shadow_pte;
-	}
-	for (i=0; i < PTRS_PER_PTE; i++) {
-		pte_clear(mm, vmaddr, pte + i);
-		vmaddr += PAGE_SIZE;
-	}
-	return pte;
-}
-
-static inline struct page *
-pte_alloc_one(struct mm_struct *mm, unsigned long vmaddr)
-{
-	pte_t *pte = pte_alloc_one_kernel(mm, vmaddr);
-	if (pte)
-		return virt_to_page(pte);
-	return NULL;
-}
-
-static inline void pte_free_kernel(pte_t *pte)
-{
-	pte_t *shadow_pte = get_shadow_pte(pte);
-
-	if (shadow_pte)
-		free_page((unsigned long) shadow_pte);
-	free_page((unsigned long) pte);
-}
-
-static inline void pte_free(struct page *pte)
-{
-	struct page *shadow_page = get_shadow_page(pte);
-
-	if (shadow_page)
-		__free_page(shadow_page);
-	__free_page(pte);
-}
-
-#define __pte_free_tlb(tlb, pte)					\
-({									\
-	struct mmu_gather *__tlb = (tlb);				\
-	struct page *__pte = (pte);					\
-	struct page *shadow_page = get_shadow_page(__pte);		\
-	if (shadow_page)						\
-		tlb_remove_page(__tlb, shadow_page);			\
-	tlb_remove_page(__tlb, __pte);					\
-})
+#define pte_alloc_one_kernel(mm, vmaddr) \
+	((pte_t *) page_table_alloc(s390_noexec))
+#define pte_alloc_one(mm, vmaddr) \
+	virt_to_page(page_table_alloc(s390_noexec))
+
+#define pte_free_kernel(pte) \
+	page_table_free((unsigned long *) pte)
+#define pte_free(pte) \
+	page_table_free((unsigned long *) page_to_phys((struct page *) pte))
 
 #endif /* _S390_PGALLOC_H */
diff --git a/include/asm-s390/pgtable.h b/include/asm-s390/pgtable.h
index 39bb5192dc3..f2cc25b74ad 100644
--- a/include/asm-s390/pgtable.h
+++ b/include/asm-s390/pgtable.h
@@ -13,8 +13,6 @@
 #ifndef _ASM_S390_PGTABLE_H
 #define _ASM_S390_PGTABLE_H
 
-#include <asm-generic/4level-fixup.h>
-
 /*
  * The Linux memory management assumes a three-level page table setup. For
  * s390 31 bit we "fold" the mid level into the top-level page table, so
@@ -35,9 +33,6 @@
 #include <asm/bug.h>
 #include <asm/processor.h>
 
-struct vm_area_struct; /* forward declaration (include/linux/mm.h) */
-struct mm_struct;
-
 extern pgd_t swapper_pg_dir[] __attribute__ ((aligned (4096)));
 extern void paging_init(void);
 extern void vmem_map_init(void);
@@ -63,14 +58,18 @@ extern char empty_zero_page[PAGE_SIZE];
  */
 #ifndef __s390x__
 # define PMD_SHIFT	22
+# define PUD_SHIFT	22
 # define PGDIR_SHIFT	22
 #else /* __s390x__ */
 # define PMD_SHIFT	21
+# define PUD_SHIFT	31
 # define PGDIR_SHIFT	31
 #endif /* __s390x__ */
 
 #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 PGDIR_SIZE      (1UL << PGDIR_SHIFT)
 #define PGDIR_MASK      (~(PGDIR_SIZE-1))
 
@@ -83,10 +82,12 @@ extern char empty_zero_page[PAGE_SIZE];
 #ifndef __s390x__
 # define PTRS_PER_PTE    1024
 # define PTRS_PER_PMD    1
+# define PTRS_PER_PUD	1
 # define PTRS_PER_PGD    512
 #else /* __s390x__ */
 # define PTRS_PER_PTE    512
 # define PTRS_PER_PMD    1024
+# define PTRS_PER_PUD	1
 # define PTRS_PER_PGD    2048
 #endif /* __s390x__ */
 
@@ -96,6 +97,8 @@ extern char empty_zero_page[PAGE_SIZE];
 	printk("%s:%d: bad pte %p.\n", __FILE__, __LINE__, (void *) pte_val(e))
 #define pmd_ERROR(e) \
 	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 pgd_ERROR(e) \
 	printk("%s:%d: bad pgd %p.\n", __FILE__, __LINE__, (void *) pgd_val(e))
 
@@ -195,7 +198,7 @@ extern unsigned long vmalloc_end;
  * I Segment-Invalid Bit:    Segment is not available for address-translation
  * TT Type 01
  * TF
- * TL Table lenght
+ * TL Table length
  *
  * The 64 bit regiontable origin of S390 has following format:
  * |      region table origon                          |       DTTL
@@ -221,6 +224,8 @@ extern unsigned long vmalloc_end;
 /* Hardware bits in the page table entry */
 #define _PAGE_RO	0x200		/* HW read-only bit  */
 #define _PAGE_INVALID	0x400		/* HW invalid bit    */
+
+/* Software bits in the page table entry */
 #define _PAGE_SWT	0x001		/* SW pte type bit t */
 #define _PAGE_SWX	0x002		/* SW pte type bit x */
 
@@ -264,60 +269,75 @@ extern unsigned long vmalloc_end;
 
 #ifndef __s390x__
 
-/* Bits in the segment table entry */
-#define _PAGE_TABLE_LEN 0xf            /* only full page-tables            */
-#define _PAGE_TABLE_COM 0x10           /* common page-table                */
-#define _PAGE_TABLE_INV 0x20           /* invalid page-table               */
-#define _SEG_PRESENT    0x001          /* Software (overlap with PTL)      */
-
-/* Bits int the storage key */
-#define _PAGE_CHANGED    0x02          /* HW changed bit                   */
-#define _PAGE_REFERENCED 0x04          /* HW referenced bit                */
-
-#define _USER_SEG_TABLE_LEN    0x7f    /* user-segment-table up to 2 GB    */
-#define _KERNEL_SEG_TABLE_LEN  0x7f    /* kernel-segment-table up to 2 GB  */
-
-/*
- * User and Kernel pagetables are identical
- */
-#define _PAGE_TABLE	_PAGE_TABLE_LEN
-#define _KERNPG_TABLE	_PAGE_TABLE_LEN
-
-/*
- * The Kernel segment-tables includes the User segment-table
- */
+/* Bits in the segment table address-space-control-element */
+#define _ASCE_SPACE_SWITCH	0x80000000UL	/* space switch event	    */
+#define _ASCE_ORIGIN_MASK	0x7ffff000UL	/* segment table origin	    */
+#define _ASCE_PRIVATE_SPACE	0x100	/* private space control	    */
+#define _ASCE_ALT_EVENT		0x80	/* storage alteration event control */
+#define _ASCE_TABLE_LENGTH	0x7f	/* 128 x 64 entries = 8k	    */
 
-#define _SEGMENT_TABLE	(_USER_SEG_TABLE_LEN|0x80000000|0x100)
-#define _KERNSEG_TABLE	_KERNEL_SEG_TABLE_LEN
+/* Bits in the segment table entry */
+#define _SEGMENT_ENTRY_ORIGIN	0x7fffffc0UL	/* page table origin	    */
+#define _SEGMENT_ENTRY_INV	0x20	/* invalid segment table entry	    */
+#define _SEGMENT_ENTRY_COMMON	0x10	/* common segment bit		    */
+#define _SEGMENT_ENTRY_PTL	0x0f	/* page table length		    */
 
-#define USER_STD_MASK	0x00000080UL
+#define _SEGMENT_ENTRY		(_SEGMENT_ENTRY_PTL)
+#define _SEGMENT_ENTRY_EMPTY	(_SEGMENT_ENTRY_INV)
 
 #else /* __s390x__ */
 
+/* Bits in the segment/region table address-space-control-element */
+#define _ASCE_ORIGIN		~0xfffUL/* segment table origin		    */
+#define _ASCE_PRIVATE_SPACE	0x100	/* private space control	    */
+#define _ASCE_ALT_EVENT		0x80	/* storage alteration event control */
+#define _ASCE_SPACE_SWITCH	0x40	/* space switch event		    */
+#define _ASCE_REAL_SPACE	0x20	/* real space control		    */
+#define _ASCE_TYPE_MASK		0x0c	/* asce table type mask		    */
+#define _ASCE_TYPE_REGION1	0x0c	/* region first table type	    */
+#define _ASCE_TYPE_REGION2	0x08	/* region second table type	    */
+#define _ASCE_TYPE_REGION3	0x04	/* region third table type	    */
+#define _ASCE_TYPE_SEGMENT	0x00	/* segment table type		    */
+#define _ASCE_TABLE_LENGTH	0x03	/* region table length		    */
+
+/* Bits in the region table entry */
+#define _REGION_ENTRY_ORIGIN	~0xfffUL/* region/segment table origin	    */
+#define _REGION_ENTRY_INV	0x20	/* invalid region table entry	    */
+#define _REGION_ENTRY_TYPE_MASK	0x0c	/* region/segment table type mask   */
+#define _REGION_ENTRY_TYPE_R1	0x0c	/* region first table type	    */
+#define _REGION_ENTRY_TYPE_R2	0x08	/* region second table type	    */
+#define _REGION_ENTRY_TYPE_R3	0x04	/* region third table type	    */
+#define _REGION_ENTRY_LENGTH	0x03	/* region third length		    */
+
+#define _REGION1_ENTRY		(_REGION_ENTRY_TYPE_R1 | _REGION_ENTRY_LENGTH)
+#define _REGION1_ENTRY_EMPTY	(_REGION_ENTRY_TYPE_R1 | _REGION_ENTRY_INV)
+#define _REGION2_ENTRY		(_REGION_ENTRY_TYPE_R2 | _REGION_ENTRY_LENGTH)
+#define _REGION2_ENTRY_EMPTY	(_REGION_ENTRY_TYPE_R2 | _REGION_ENTRY_INV)
+#define _REGION3_ENTRY		(_REGION_ENTRY_TYPE_R3 | _REGION_ENTRY_LENGTH)
+#define _REGION3_ENTRY_EMPTY	(_REGION_ENTRY_TYPE_R3 | _REGION_ENTRY_INV)
+
 /* Bits in the segment table entry */
-#define _PMD_ENTRY_INV   0x20          /* invalid segment table entry      */
-#define _PMD_ENTRY       0x00        
+#define _SEGMENT_ENTRY_ORIGIN	~0x7ffUL/* segment table origin		    */
+#define _SEGMENT_ENTRY_RO	0x200	/* page protection bit		    */
+#define _SEGMENT_ENTRY_INV	0x20	/* invalid segment table entry	    */
+
+#define _SEGMENT_ENTRY		(0)
+#define _SEGMENT_ENTRY_EMPTY	(_SEGMENT_ENTRY_INV)
 
-/* Bits in the region third table entry */
-#define _PGD_ENTRY_INV   0x20          /* invalid region table entry       */
-#define _PGD_ENTRY       0x07
+#endif /* __s390x__ */
 
 /*
- * User and kernel page directory
+ * A user page table pointer has the space-switch-event bit, the
+ * private-space-control bit and the storage-alteration-event-control
+ * bit set. A kernel page table pointer doesn't need them.
  */
-#define _REGION_THIRD       0x4
-#define _REGION_THIRD_LEN   0x3 
-#define _REGION_TABLE       (_REGION_THIRD|_REGION_THIRD_LEN|0x40|0x100)
-#define _KERN_REGION_TABLE  (_REGION_THIRD|_REGION_THIRD_LEN)
-
-#define USER_STD_MASK           0x0000000000000080UL
+#define _ASCE_USER_BITS		(_ASCE_SPACE_SWITCH | _ASCE_PRIVATE_SPACE | \
+				 _ASCE_ALT_EVENT)
 
-/* Bits in the storage key */
+/* Bits int the storage key */
 #define _PAGE_CHANGED    0x02          /* HW changed bit                   */
 #define _PAGE_REFERENCED 0x04          /* HW referenced bit                */
 
-#endif /* __s390x__ */
-
 /*
  * Page protection definitions.
  */
@@ -358,65 +378,38 @@ extern unsigned long vmalloc_end;
 #define __S111	PAGE_EX_RW
 
 #ifndef __s390x__
-# define PMD_SHADOW_SHIFT	1
-# define PGD_SHADOW_SHIFT	1
+# define PxD_SHADOW_SHIFT	1
 #else /* __s390x__ */
-# define PMD_SHADOW_SHIFT	2
-# define PGD_SHADOW_SHIFT	2
+# define PxD_SHADOW_SHIFT	2
 #endif /* __s390x__ */
 
 static inline struct page *get_shadow_page(struct page *page)
 {
-	if (s390_noexec && !list_empty(&page->lru))
-		return virt_to_page(page->lru.next);
-	return NULL;
-}
-
-static inline pte_t *get_shadow_pte(pte_t *ptep)
-{
-	unsigned long pteptr = (unsigned long) (ptep);
-
-	if (s390_noexec) {
-		unsigned long offset = pteptr & (PAGE_SIZE - 1);
-		void *addr = (void *) (pteptr ^ offset);
-		struct page *page = virt_to_page(addr);
-		if (!list_empty(&page->lru))
-			return (pte_t *) ((unsigned long) page->lru.next |
-								offset);
-	}
+	if (s390_noexec && page->index)
+		return virt_to_page((void *)(addr_t) page->index);
 	return NULL;
 }
 
-static inline pmd_t *get_shadow_pmd(pmd_t *pmdp)
+static inline void *get_shadow_pte(void *table)
 {
-	unsigned long pmdptr = (unsigned long) (pmdp);
+	unsigned long addr, offset;
+	struct page *page;
 
-	if (s390_noexec) {
-		unsigned long offset = pmdptr &
-				((PAGE_SIZE << PMD_SHADOW_SHIFT) - 1);
-		void *addr = (void *) (pmdptr ^ offset);
-		struct page *page = virt_to_page(addr);
-		if (!list_empty(&page->lru))
-			return (pmd_t *) ((unsigned long) page->lru.next |
-								offset);
-	}
-	return NULL;
+	addr = (unsigned long) table;
+	offset = addr & (PAGE_SIZE - 1);
+	page = virt_to_page((void *)(addr ^ offset));
+	return (void *)(addr_t)(page->index ? (page->index | offset) : 0UL);
 }
 
-static inline pgd_t *get_shadow_pgd(pgd_t *pgdp)
+static inline void *get_shadow_table(void *table)
 {
-	unsigned long pgdptr = (unsigned long) (pgdp);
+	unsigned long addr, offset;
+	struct page *page;
 
-	if (s390_noexec) {
-		unsigned long offset = pgdptr &
-				((PAGE_SIZE << PGD_SHADOW_SHIFT) - 1);
-		void *addr = (void *) (pgdptr ^ offset);
-		struct page *page = virt_to_page(addr);
-		if (!list_empty(&page->lru))
-			return (pgd_t *) ((unsigned long) page->lru.next |
-								offset);
-	}
-	return NULL;
+	addr = (unsigned long) table;
+	offset = addr & ((PAGE_SIZE << PxD_SHADOW_SHIFT) - 1);
+	page = virt_to_page((void *)(addr ^ offset));
+	return (void *)(addr_t)(page->index ? (page->index | offset) : 0UL);
 }
 
 /*
@@ -424,7 +417,8 @@ static inline pgd_t *get_shadow_pgd(pgd_t *pgdp)
  * within a page table are directly modified.  Thus, the following
  * hook is made available.
  */
-static inline void set_pte(pte_t *pteptr, pte_t pteval)
+static inline void set_pte_at(struct mm_struct *mm, unsigned long addr,
+			      pte_t *pteptr, pte_t pteval)
 {
 	pte_t *shadow_pte = get_shadow_pte(pteptr);
 
@@ -437,7 +431,6 @@ static inline void set_pte(pte_t *pteptr, pte_t pteval)
 			pte_val(*shadow_pte) = _PAGE_TYPE_EMPTY;
 	}
 }
-#define set_pte_at(mm,addr,ptep,pteval) set_pte(ptep,pteval)
 
 /*
  * pgd/pmd/pte query functions
@@ -448,47 +441,50 @@ static inline int pgd_present(pgd_t pgd) { return 1; }
 static inline int pgd_none(pgd_t pgd)    { return 0; }
 static inline int pgd_bad(pgd_t pgd)     { return 0; }
 
-static inline int pmd_present(pmd_t pmd) { return pmd_val(pmd) & _SEG_PRESENT; }
-static inline int pmd_none(pmd_t pmd)    { return pmd_val(pmd) & _PAGE_TABLE_INV; }
-static inline int pmd_bad(pmd_t pmd)
-{
-	return (pmd_val(pmd) & (~PAGE_MASK & ~_PAGE_TABLE_INV)) != _PAGE_TABLE;
-}
+static inline int pud_present(pud_t pud) { return 1; }
+static inline int pud_none(pud_t pud)	 { return 0; }
+static inline int pud_bad(pud_t pud)	 { return 0; }
 
 #else /* __s390x__ */
 
-static inline int pgd_present(pgd_t pgd)
+static inline int pgd_present(pgd_t pgd) { return 1; }
+static inline int pgd_none(pgd_t pgd)	 { return 0; }
+static inline int pgd_bad(pgd_t pgd)	 { return 0; }
+
+static inline int pud_present(pud_t pud)
 {
-	return (pgd_val(pgd) & ~PAGE_MASK) == _PGD_ENTRY;
+	return pud_val(pud) & _REGION_ENTRY_ORIGIN;
 }
 
-static inline int pgd_none(pgd_t pgd)
+static inline int pud_none(pud_t pud)
 {
-	return pgd_val(pgd) & _PGD_ENTRY_INV;
+	return pud_val(pud) & _REGION_ENTRY_INV;
 }
 
-static inline int pgd_bad(pgd_t pgd)
+static inline int pud_bad(pud_t pud)
 {
-	return (pgd_val(pgd) & (~PAGE_MASK & ~_PGD_ENTRY_INV)) != _PGD_ENTRY;
+	unsigned long mask = ~_REGION_ENTRY_ORIGIN & ~_REGION_ENTRY_INV;
+	return (pud_val(pud) & mask) != _REGION3_ENTRY;
 }
 
+#endif /* __s390x__ */
+
 static inline int pmd_present(pmd_t pmd)
 {
-	return (pmd_val(pmd) & ~PAGE_MASK) == _PMD_ENTRY;
+	return pmd_val(pmd) & _SEGMENT_ENTRY_ORIGIN;
 }
 
 static inline int pmd_none(pmd_t pmd)
 {
-	return pmd_val(pmd) & _PMD_ENTRY_INV;
+	return pmd_val(pmd) & _SEGMENT_ENTRY_INV;
 }
 
 static inline int pmd_bad(pmd_t pmd)
 {
-	return (pmd_val(pmd) & (~PAGE_MASK & ~_PMD_ENTRY_INV)) != _PMD_ENTRY;
+	unsigned long mask = ~_SEGMENT_ENTRY_ORIGIN & ~_SEGMENT_ENTRY_INV;
+	return (pmd_val(pmd) & mask) != _SEGMENT_ENTRY;
 }
 
-#endif /* __s390x__ */
-
 static inline int pte_none(pte_t pte)
 {
 	return (pte_val(pte) & _PAGE_INVALID) && !(pte_val(pte) & _PAGE_SWT);
@@ -508,7 +504,8 @@ static inline int pte_file(pte_t pte)
 	return (pte_val(pte) & mask) == _PAGE_TYPE_FILE;
 }
 
-#define pte_same(a,b)	(pte_val(a) == pte_val(b))
+#define __HAVE_ARCH_PTE_SAME
+#define pte_same(a,b)  (pte_val(a) == pte_val(b))
 
 /*
  * query functions pte_write/pte_dirty/pte_young only work if
@@ -543,58 +540,52 @@ static inline int pte_young(pte_t pte)
 
 #ifndef __s390x__
 
-static inline void pgd_clear(pgd_t * pgdp)      { }
+#define pgd_clear(pgd)		do { } while (0)
+#define pud_clear(pud)		do { } while (0)
 
 static inline void pmd_clear_kernel(pmd_t * pmdp)
 {
-	pmd_val(pmdp[0]) = _PAGE_TABLE_INV;
-	pmd_val(pmdp[1]) = _PAGE_TABLE_INV;
-	pmd_val(pmdp[2]) = _PAGE_TABLE_INV;
-	pmd_val(pmdp[3]) = _PAGE_TABLE_INV;
-}
-
-static inline void pmd_clear(pmd_t * pmdp)
-{
-	pmd_t *shadow_pmd = get_shadow_pmd(pmdp);
-
-	pmd_clear_kernel(pmdp);
-	if (shadow_pmd)
-		pmd_clear_kernel(shadow_pmd);
+	pmd_val(pmdp[0]) = _SEGMENT_ENTRY_EMPTY;
+	pmd_val(pmdp[1]) = _SEGMENT_ENTRY_EMPTY;
+	pmd_val(pmdp[2]) = _SEGMENT_ENTRY_EMPTY;
+	pmd_val(pmdp[3]) = _SEGMENT_ENTRY_EMPTY;
 }
 
 #else /* __s390x__ */
 
-static inline void pgd_clear_kernel(pgd_t * pgdp)
+#define pgd_clear(pgd)		do { } while (0)
+
+static inline void pud_clear_kernel(pud_t *pud)
 {
-	pgd_val(*pgdp) = _PGD_ENTRY_INV | _PGD_ENTRY;
+	pud_val(*pud) = _REGION3_ENTRY_EMPTY;
 }
 
-static inline void pgd_clear(pgd_t * pgdp)
+static inline void pud_clear(pud_t * pud)
 {
-	pgd_t *shadow_pgd = get_shadow_pgd(pgdp);
+	pud_t *shadow = get_shadow_table(pud);
 
-	pgd_clear_kernel(pgdp);
-	if (shadow_pgd)
-		pgd_clear_kernel(shadow_pgd);
+	pud_clear_kernel(pud);
+	if (shadow)
+		pud_clear_kernel(shadow);
 }
 
 static inline void pmd_clear_kernel(pmd_t * pmdp)
 {
-	pmd_val(*pmdp) = _PMD_ENTRY_INV | _PMD_ENTRY;
-	pmd_val1(*pmdp) = _PMD_ENTRY_INV | _PMD_ENTRY;
+	pmd_val(*pmdp) = _SEGMENT_ENTRY_EMPTY;
+	pmd_val1(*pmdp) = _SEGMENT_ENTRY_EMPTY;
 }
 
+#endif /* __s390x__ */
+
 static inline void pmd_clear(pmd_t * pmdp)
 {
-	pmd_t *shadow_pmd = get_shadow_pmd(pmdp);
+	pmd_t *shadow_pmd = get_shadow_table(pmdp);
 
 	pmd_clear_kernel(pmdp);
 	if (shadow_pmd)
 		pmd_clear_kernel(shadow_pmd);
 }
 
-#endif /* __s390x__ */
-
 static inline void pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
 {
 	pte_t *shadow_pte = get_shadow_pte(ptep);
@@ -663,24 +654,19 @@ static inline pte_t pte_mkyoung(pte_t pte)
 	return pte;
 }
 
-static inline int ptep_test_and_clear_young(struct vm_area_struct *vma, unsigned long addr, pte_t *ptep)
+#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
+static inline int ptep_test_and_clear_young(struct vm_area_struct *vma,
+					    unsigned long addr, pte_t *ptep)
 {
 	return 0;
 }
 
-static inline int
-ptep_clear_flush_young(struct vm_area_struct *vma,
-			unsigned long address, pte_t *ptep)
+#define __HAVE_ARCH_PTEP_CLEAR_YOUNG_FLUSH
+static inline int ptep_clear_flush_young(struct vm_area_struct *vma,
+					 unsigned long address, pte_t *ptep)
 {
 	/* No need to flush TLB; bits are in storage key */
-	return ptep_test_and_clear_young(vma, address, ptep);
-}
-
-static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
-{
-	pte_t pte = *ptep;
-	pte_clear(mm, addr, ptep);
-	return pte;
+	return 0;
 }
 
 static inline void __ptep_ipte(unsigned long address, pte_t *ptep)
@@ -709,6 +695,32 @@ static inline void ptep_invalidate(unsigned long address, pte_t *ptep)
 		__ptep_ipte(address, ptep);
 }
 
+/*
+ * This is hard to understand. ptep_get_and_clear and ptep_clear_flush
+ * both clear the TLB for the unmapped pte. The reason is that
+ * ptep_get_and_clear is used in common code (e.g. change_pte_range)
+ * to modify an active pte. The sequence is
+ *   1) ptep_get_and_clear
+ *   2) set_pte_at
+ *   3) flush_tlb_range
+ * On s390 the tlb needs to get flushed with the modification of the pte
+ * if the pte is active. The only way how this can be implemented is to
+ * have ptep_get_and_clear do the tlb flush. In exchange flush_tlb_range
+ * is a nop.
+ */
+#define __HAVE_ARCH_PTEP_GET_AND_CLEAR
+#define ptep_get_and_clear(__mm, __address, __ptep)			\
+({									\
+	pte_t __pte = *(__ptep);					\
+	if (atomic_read(&(__mm)->mm_users) > 1 ||			\
+	    (__mm) != current->active_mm)				\
+		ptep_invalidate(__address, __ptep);			\
+	else								\
+		pte_clear((__mm), (__address), (__ptep));		\
+	__pte;								\
+})
+
+#define __HAVE_ARCH_PTEP_CLEAR_FLUSH
 static inline pte_t ptep_clear_flush(struct vm_area_struct *vma,
 				     unsigned long address, pte_t *ptep)
 {
@@ -717,12 +729,40 @@ static inline pte_t ptep_clear_flush(struct vm_area_struct *vma,
 	return pte;
 }
 
-static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
+/*
+ * The batched pte unmap code uses ptep_get_and_clear_full to clear the
+ * ptes. Here an optimization is possible. tlb_gather_mmu flushes all
+ * tlbs of an mm if it can guarantee that the ptes of the mm_struct
+ * cannot be accessed while the batched unmap is running. In this case
+ * full==1 and a simple pte_clear is enough. See tlb.h.
+ */
+#define __HAVE_ARCH_PTEP_GET_AND_CLEAR_FULL
+static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm,
+					    unsigned long addr,
+					    pte_t *ptep, int full)
 {
-	pte_t old_pte = *ptep;
-	set_pte_at(mm, addr, ptep, pte_wrprotect(old_pte));
+	pte_t pte = *ptep;
+
+	if (full)
+		pte_clear(mm, addr, ptep);
+	else
+		ptep_invalidate(addr, ptep);
+	return pte;
 }
 
+#define __HAVE_ARCH_PTEP_SET_WRPROTECT
+#define ptep_set_wrprotect(__mm, __addr, __ptep)			\
+({									\
+	pte_t __pte = *(__ptep);					\
+	if (pte_write(__pte)) {						\
+		if (atomic_read(&(__mm)->mm_users) > 1 ||		\
+		    (__mm) != current->active_mm)			\
+			ptep_invalidate(__addr, __ptep);		\
+		set_pte_at(__mm, __addr, __ptep, pte_wrprotect(__pte));	\
+	}								\
+})
+
+#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
 #define ptep_set_access_flags(__vma, __addr, __ptep, __entry, __dirty)	\
 ({									\
 	int __changed = !pte_same(*(__ptep), __entry);			\
@@ -740,11 +780,13 @@ static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr,
  * should therefore only be called if it is not mapped in any
  * address space.
  */
+#define __HAVE_ARCH_PAGE_TEST_DIRTY
 static inline int page_test_dirty(struct page *page)
 {
 	return (page_get_storage_key(page_to_phys(page)) & _PAGE_CHANGED) != 0;
 }
 
+#define __HAVE_ARCH_PAGE_CLEAR_DIRTY
 static inline void page_clear_dirty(struct page *page)
 {
 	page_set_storage_key(page_to_phys(page), PAGE_DEFAULT_KEY);
@@ -753,6 +795,7 @@ static inline void page_clear_dirty(struct page *page)
 /*
  * Test and clear referenced bit in storage key.
  */
+#define __HAVE_ARCH_PAGE_TEST_AND_CLEAR_YOUNG
 static inline int page_test_and_clear_young(struct page *page)
 {
 	unsigned long physpage = page_to_phys(page);
@@ -784,63 +827,48 @@ static inline pte_t mk_pte(struct page *page, pgprot_t pgprot)
 	return mk_pte_phys(physpage, pgprot);
 }
 
-static inline pte_t pfn_pte(unsigned long pfn, pgprot_t pgprot)
-{
-	unsigned long physpage = __pa((pfn) << PAGE_SHIFT);
-
-	return mk_pte_phys(physpage, pgprot);
-}
-
-#ifdef __s390x__
-
-static inline pmd_t pfn_pmd(unsigned long pfn, pgprot_t pgprot)
-{
-	unsigned long physpage = __pa((pfn) << PAGE_SHIFT);
-
-	return __pmd(physpage + pgprot_val(pgprot));
-}
-
-#endif /* __s390x__ */
-
-#define pte_pfn(x) (pte_val(x) >> PAGE_SHIFT)
-#define pte_page(x) pfn_to_page(pte_pfn(x))
+#define pgd_index(address) (((address) >> PGDIR_SHIFT) & (PTRS_PER_PGD-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))
 
-#define pmd_page_vaddr(pmd) (pmd_val(pmd) & PAGE_MASK)
+#define pgd_offset(mm, address) ((mm)->pgd + pgd_index(address))
+#define pgd_offset_k(address) pgd_offset(&init_mm, address)
 
-#define pmd_page(pmd) pfn_to_page(pmd_val(pmd) >> PAGE_SHIFT)
+#ifndef __s390x__
 
-#define pgd_page_vaddr(pgd) (pgd_val(pgd) & PAGE_MASK)
+#define pmd_deref(pmd) (pmd_val(pmd) & _SEGMENT_ENTRY_ORIGIN)
+#define pud_deref(pmd) ({ BUG(); 0UL; })
+#define pgd_deref(pmd) ({ BUG(); 0UL; })
 
-#define pgd_page(pgd) pfn_to_page(pgd_val(pgd) >> PAGE_SHIFT)
+#define pud_offset(pgd, address) ((pud_t *) pgd)
+#define pmd_offset(pud, address) ((pmd_t *) pud + pmd_index(address))
 
-/* to find an entry in a page-table-directory */
-#define pgd_index(address) (((address) >> PGDIR_SHIFT) & (PTRS_PER_PGD-1))
-#define pgd_offset(mm, address) ((mm)->pgd+pgd_index(address))
+#else /* __s390x__ */
 
-/* to find an entry in a kernel page-table-directory */
-#define pgd_offset_k(address) pgd_offset(&init_mm, address)
+#define pmd_deref(pmd) (pmd_val(pmd) & _SEGMENT_ENTRY_ORIGIN)
+#define pud_deref(pud) (pud_val(pud) & _REGION_ENTRY_ORIGIN)
+#define pgd_deref(pgd) ({ BUG(); 0UL; })
 
-#ifndef __s390x__
+#define pud_offset(pgd, address) ((pud_t *) pgd)
 
-/* Find an entry in the second-level page table.. */
-static inline pmd_t * pmd_offset(pgd_t * dir, unsigned long address)
+static inline pmd_t *pmd_offset(pud_t *pud, unsigned long address)
 {
-        return (pmd_t *) dir;
+	pmd_t *pmd = (pmd_t *) pud_deref(*pud);
+	return pmd + pmd_index(address);
 }
 
-#else /* __s390x__ */
+#endif /* __s390x__ */
 
-/* Find an entry in the second-level page table.. */
-#define pmd_index(address) (((address) >> PMD_SHIFT) & (PTRS_PER_PMD-1))
-#define pmd_offset(dir,addr) \
-	((pmd_t *) pgd_page_vaddr(*(dir)) + pmd_index(addr))
+#define pfn_pte(pfn,pgprot) mk_pte_phys(__pa((pfn) << PAGE_SHIFT),(pgprot))
+#define pte_pfn(x) (pte_val(x) >> PAGE_SHIFT)
+#define pte_page(x) pfn_to_page(pte_pfn(x))
 
-#endif /* __s390x__ */
+#define pmd_page(pmd) pfn_to_page(pmd_val(pmd) >> PAGE_SHIFT)
 
-/* Find an entry in the third-level page table.. */
-#define pte_index(address) (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE-1))
-#define pte_offset_kernel(pmd, address) \
-	((pte_t *) pmd_page_vaddr(*(pmd)) + pte_index(address))
+/* Find an entry in the lowest level page table.. */
+#define pte_offset(pmd, addr) ((pte_t *) pmd_deref(*(pmd)) + pte_index(addr))
+#define pte_offset_kernel(pmd, address) pte_offset(pmd,address)
 #define pte_offset_map(pmd, address) pte_offset_kernel(pmd, address)
 #define pte_offset_map_nested(pmd, address) pte_offset_kernel(pmd, address)
 #define pte_unmap(pte) do { } while (0)
@@ -930,17 +958,6 @@ extern int remove_shared_memory(unsigned long start, unsigned long size);
 #define __HAVE_ARCH_MEMMAP_INIT
 extern void memmap_init(unsigned long, int, unsigned long, unsigned long);
 
-#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
-#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
-#define __HAVE_ARCH_PTEP_CLEAR_YOUNG_FLUSH
-#define __HAVE_ARCH_PTEP_GET_AND_CLEAR
-#define __HAVE_ARCH_PTEP_CLEAR_FLUSH
-#define __HAVE_ARCH_PTEP_SET_WRPROTECT
-#define __HAVE_ARCH_PTE_SAME
-#define __HAVE_ARCH_PAGE_TEST_DIRTY
-#define __HAVE_ARCH_PAGE_CLEAR_DIRTY
-#define __HAVE_ARCH_PAGE_TEST_AND_CLEAR_YOUNG
 #include <asm-generic/pgtable.h>
 
 #endif /* _S390_PAGE_H */
-
diff --git a/include/asm-s390/processor.h b/include/asm-s390/processor.h
index 3b972d4c6b2..21d40a19355 100644
--- a/include/asm-s390/processor.h
+++ b/include/asm-s390/processor.h
@@ -93,7 +93,6 @@ struct thread_struct {
 	s390_fp_regs fp_regs;
 	unsigned int  acrs[NUM_ACRS];
         unsigned long ksp;              /* kernel stack pointer             */
-        unsigned long user_seg;         /* HSTD                             */
 	mm_segment_t mm_segment;
         unsigned long prot_addr;        /* address of protection-excep.     */
         unsigned int error_code;        /* error-code of last prog-excep.   */
@@ -128,22 +127,9 @@ struct stack_frame {
 
 #define ARCH_MIN_TASKALIGN	8
 
-#ifndef __s390x__
-# define __SWAPPER_PG_DIR __pa(&swapper_pg_dir[0]) + _SEGMENT_TABLE
-#else /* __s390x__ */
-# define __SWAPPER_PG_DIR __pa(&swapper_pg_dir[0]) + _REGION_TABLE
-#endif /* __s390x__ */
-
-#define INIT_THREAD {{0,{{0},{0},{0},{0},{0},{0},{0},{0},{0},{0},	       \
-			    {0},{0},{0},{0},{0},{0}}},			       \
-		     {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},	       \
-		     sizeof(init_stack) + (unsigned long) &init_stack,	       \
-		     __SWAPPER_PG_DIR,					       \
-		     {0},						       \
-		     0,0,0,						       \
-		     (per_struct) {{{{0,}}},0,0,0,0,{{0,}}},		       \
-		     0, 0						       \
-} 
+#define INIT_THREAD {							\
+	.ksp = sizeof(init_stack) + (unsigned long) &init_stack,	\
+}
 
 /*
  * Do necessary setup to start up a new thread.
diff --git a/include/asm-s390/tlb.h b/include/asm-s390/tlb.h
index 51bd957b85b..618693cfc10 100644
--- a/include/asm-s390/tlb.h
+++ b/include/asm-s390/tlb.h
@@ -2,19 +2,130 @@
 #define _S390_TLB_H
 
 /*
- * s390 doesn't need any special per-pte or
- * per-vma handling..
+ * TLB flushing on s390 is complicated. The following requirement
+ * from the principles of operation is the most arduous:
+ *
+ * "A valid table entry must not be changed while it is attached
+ * to any CPU and may be used for translation by that CPU except to
+ * (1) invalidate the entry by using INVALIDATE PAGE TABLE ENTRY,
+ * or INVALIDATE DAT TABLE ENTRY, (2) alter bits 56-63 of a page
+ * table entry, or (3) make a change by means of a COMPARE AND SWAP
+ * AND PURGE instruction that purges the TLB."
+ *
+ * The modification of a pte of an active mm struct therefore is
+ * a two step process: i) invalidate the pte, ii) store the new pte.
+ * This is true for the page protection bit as well.
+ * The only possible optimization is to flush at the beginning of
+ * a tlb_gather_mmu cycle if the mm_struct is currently not in use.
+ *
+ * Pages used for the page tables is a different story. FIXME: more
  */
-#define tlb_start_vma(tlb, vma) do { } while (0)
-#define tlb_end_vma(tlb, vma) do { } while (0)
-#define __tlb_remove_tlb_entry(tlb, ptep, address) do { } while (0)
+
+#include <linux/mm.h>
+#include <linux/swap.h>
+#include <asm/processor.h>
+#include <asm/pgalloc.h>
+#include <asm/smp.h>
+#include <asm/tlbflush.h>
+
+#ifndef CONFIG_SMP
+#define TLB_NR_PTRS	1
+#else
+#define TLB_NR_PTRS	508
+#endif
+
+struct mmu_gather {
+	struct mm_struct *mm;
+	unsigned int fullmm;
+	unsigned int nr_ptes;
+	unsigned int nr_pmds;
+	void *array[TLB_NR_PTRS];
+};
+
+DECLARE_PER_CPU(struct mmu_gather, mmu_gathers);
+
+static inline struct mmu_gather *tlb_gather_mmu(struct mm_struct *mm,
+						unsigned int full_mm_flush)
+{
+	struct mmu_gather *tlb = &get_cpu_var(mmu_gathers);
+
+	tlb->mm = mm;
+	tlb->fullmm = full_mm_flush || (num_online_cpus() == 1) ||
+		(atomic_read(&mm->mm_users) <= 1 && mm == current->active_mm);
+	tlb->nr_ptes = 0;
+	tlb->nr_pmds = TLB_NR_PTRS;
+	if (tlb->fullmm)
+		__tlb_flush_mm(mm);
+	return tlb;
+}
+
+static inline void tlb_flush_mmu(struct mmu_gather *tlb,
+				 unsigned long start, unsigned long end)
+{
+	if (!tlb->fullmm && (tlb->nr_ptes > 0 || tlb->nr_pmds < TLB_NR_PTRS))
+		__tlb_flush_mm(tlb->mm);
+	while (tlb->nr_ptes > 0)
+		pte_free(tlb->array[--tlb->nr_ptes]);
+	while (tlb->nr_pmds < TLB_NR_PTRS)
+		pmd_free((pmd_t *) tlb->array[tlb->nr_pmds++]);
+}
+
+static inline void tlb_finish_mmu(struct mmu_gather *tlb,
+				  unsigned long start, unsigned long end)
+{
+	tlb_flush_mmu(tlb, start, end);
+
+	/* keep the page table cache within bounds */
+	check_pgt_cache();
+
+	put_cpu_var(mmu_gathers);
+}
 
 /*
- * .. because we flush the whole mm when it
- * fills up.
+ * Release the page cache reference for a pte removed by
+ * tlb_ptep_clear_flush. In both flush modes the tlb fo a page cache page
+ * has already been freed, so just do free_page_and_swap_cache.
  */
-#define tlb_flush(tlb) flush_tlb_mm((tlb)->mm)
+static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page)
+{
+	free_page_and_swap_cache(page);
+}
 
-#include <asm-generic/tlb.h>
+/*
+ * pte_free_tlb frees a pte table and clears the CRSTE for the
+ * page table from the tlb.
+ */
+static inline void pte_free_tlb(struct mmu_gather *tlb, struct page *page)
+{
+	if (!tlb->fullmm) {
+		tlb->array[tlb->nr_ptes++] = page;
+		if (tlb->nr_ptes >= tlb->nr_pmds)
+			tlb_flush_mmu(tlb, 0, 0);
+	} else
+		pte_free(page);
+}
 
+/*
+ * pmd_free_tlb frees a pmd table and clears the CRSTE for the
+ * segment table entry from the tlb.
+ */
+static inline void pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd)
+{
+#ifdef __s390x__
+	if (!tlb->fullmm) {
+		tlb->array[--tlb->nr_pmds] = (struct page *) pmd;
+		if (tlb->nr_ptes >= tlb->nr_pmds)
+			tlb_flush_mmu(tlb, 0, 0);
+	} else
+		pmd_free(pmd);
 #endif
+}
+
+#define pud_free_tlb(tlb, pud)			do { } while (0)
+
+#define tlb_start_vma(tlb, vma)			do { } while (0)
+#define tlb_end_vma(tlb, vma)			do { } while (0)
+#define tlb_remove_tlb_entry(tlb, ptep, addr)	do { } while (0)
+#define tlb_migrate_finish(mm)			do { } while (0)
+
+#endif /* _S390_TLB_H */
diff --git a/include/asm-s390/tlbflush.h b/include/asm-s390/tlbflush.h
index 6de2632a3e4..a69bd2490d5 100644
--- a/include/asm-s390/tlbflush.h
+++ b/include/asm-s390/tlbflush.h
@@ -6,68 +6,19 @@
 #include <asm/pgalloc.h>
 
 /*
- * TLB flushing:
- *
- *  - flush_tlb() flushes the current mm struct TLBs
- *  - flush_tlb_all() flushes all processes TLBs 
- *  - flush_tlb_mm(mm) flushes the specified mm context TLB's
- *  - flush_tlb_page(vma, vmaddr) flushes one page
- *  - flush_tlb_range(vma, start, end) flushes a range of pages
- *  - flush_tlb_kernel_range(start, end) flushes a range of kernel pages
- */
-
-/*
- * S/390 has three ways of flushing TLBs
- * 'ptlb' does a flush of the local processor
- * 'csp' flushes the TLBs on all PUs of a SMP
- * 'ipte' invalidates a pte in a page table and flushes that out of
- * the TLBs of all PUs of a SMP
- */
-
-#define local_flush_tlb() \
-do {  asm volatile("ptlb": : :"memory"); } while (0)
-
-#ifndef CONFIG_SMP
-
-/*
- * We always need to flush, since s390 does not flush tlb
- * on each context switch
+ * Flush all tlb entries on the local cpu.
  */
-
-static inline void flush_tlb(void)
+static inline void __tlb_flush_local(void)
 {
-	local_flush_tlb();
+	asm volatile("ptlb" : : : "memory");
 }
-static inline void flush_tlb_all(void)
-{
-	local_flush_tlb();
-}
-static inline void flush_tlb_mm(struct mm_struct *mm) 
-{
-	local_flush_tlb();
-}
-static inline void flush_tlb_page(struct vm_area_struct *vma,
-				  unsigned long addr)
-{
-	local_flush_tlb();
-}
-static inline void flush_tlb_range(struct vm_area_struct *vma,
-				   unsigned long start, unsigned long end)
-{
-	local_flush_tlb();
-}
-
-#define flush_tlb_kernel_range(start, end) \
-	local_flush_tlb();
-
-#else
 
-#include <asm/smp.h>
-
-extern void smp_ptlb_all(void);
-
-static inline void global_flush_tlb(void)
+/*
+ * Flush all tlb entries on all cpus.
+ */
+static inline void __tlb_flush_global(void)
 {
+	extern void smp_ptlb_all(void);
 	register unsigned long reg2 asm("2");
 	register unsigned long reg3 asm("3");
 	register unsigned long reg4 asm("4");
@@ -89,66 +40,75 @@ static inline void global_flush_tlb(void)
 }
 
 /*
- * We only have to do global flush of tlb if process run since last
- * flush on any other pu than current. 
- * If we have threads (mm->count > 1) we always do a global flush, 
- * since the process runs on more than one processor at the same time.
+ * Flush all tlb entries of a page table on all cpus.
  */
+static inline void __tlb_flush_idte(pgd_t *pgd)
+{
+	asm volatile(
+		"	.insn	rrf,0xb98e0000,0,%0,%1,0"
+		: : "a" (2048), "a" (__pa(pgd) & PAGE_MASK) : "cc" );
+}
 
-static inline void __flush_tlb_mm(struct mm_struct * mm)
+static inline void __tlb_flush_mm(struct mm_struct * mm)
 {
 	cpumask_t local_cpumask;
 
 	if (unlikely(cpus_empty(mm->cpu_vm_mask)))
 		return;
+	/*
+	 * If the machine has IDTE we prefer to do a per mm flush
+	 * on all cpus instead of doing a local flush if the mm
+	 * only ran on the local cpu.
+	 */
 	if (MACHINE_HAS_IDTE) {
-		pgd_t *shadow_pgd = get_shadow_pgd(mm->pgd);
+		pgd_t *shadow_pgd = get_shadow_table(mm->pgd);
 
-		if (shadow_pgd) {
-			asm volatile(
-				"	.insn	rrf,0xb98e0000,0,%0,%1,0"
-				: : "a" (2048),
-				"a" (__pa(shadow_pgd) & PAGE_MASK) : "cc" );
-		}
-		asm volatile(
-			"	.insn	rrf,0xb98e0000,0,%0,%1,0"
-			: : "a" (2048), "a" (__pa(mm->pgd)&PAGE_MASK) : "cc");
+		if (shadow_pgd)
+			__tlb_flush_idte(shadow_pgd);
+		__tlb_flush_idte(mm->pgd);
 		return;
 	}
 	preempt_disable();
+	/*
+	 * If the process only ran on the local cpu, do a local flush.
+	 */
 	local_cpumask = cpumask_of_cpu(smp_processor_id());
 	if (cpus_equal(mm->cpu_vm_mask, local_cpumask))
-		local_flush_tlb();
+		__tlb_flush_local();
 	else
-		global_flush_tlb();
+		__tlb_flush_global();
 	preempt_enable();
 }
 
-static inline void flush_tlb(void)
-{
-	__flush_tlb_mm(current->mm);
-}
-static inline void flush_tlb_all(void)
-{
-	global_flush_tlb();
-}
-static inline void flush_tlb_mm(struct mm_struct *mm) 
-{
-	__flush_tlb_mm(mm); 
-}
-static inline void flush_tlb_page(struct vm_area_struct *vma,
-				  unsigned long addr)
-{
-	__flush_tlb_mm(vma->vm_mm);
-}
-static inline void flush_tlb_range(struct vm_area_struct *vma,
-				   unsigned long start, unsigned long end)
+static inline void __tlb_flush_mm_cond(struct mm_struct * mm)
 {
-	__flush_tlb_mm(vma->vm_mm); 
+	if (atomic_read(&mm->mm_users) <= 1 && mm == current->active_mm)
+		__tlb_flush_mm(mm);
 }
 
-#define flush_tlb_kernel_range(start, end) global_flush_tlb()
+/*
+ * TLB flushing:
+ *  flush_tlb() - flushes the current mm struct TLBs
+ *  flush_tlb_all() - flushes all processes TLBs
+ *  flush_tlb_mm(mm) - flushes the specified mm context TLB's
+ *  flush_tlb_page(vma, vmaddr) - flushes one page
+ *  flush_tlb_range(vma, start, end) - flushes a range of pages
+ *  flush_tlb_kernel_range(start, end) - flushes a range of kernel pages
+ */
 
-#endif
+/*
+ * flush_tlb_mm goes together with ptep_set_wrprotect for the
+ * copy_page_range operation and flush_tlb_range is related to
+ * ptep_get_and_clear for change_protection. ptep_set_wrprotect and
+ * ptep_get_and_clear do not flush the TLBs directly if the mm has
+ * only one user. At the end of the update the flush_tlb_mm and
+ * flush_tlb_range functions need to do the flush.
+ */
+#define flush_tlb()				do { } while (0)
+#define flush_tlb_all()				do { } while (0)
+#define flush_tlb_mm(mm)			__tlb_flush_mm_cond(mm)
+#define flush_tlb_page(vma, addr)		do { } while (0)
+#define flush_tlb_range(vma, start, end)	__tlb_flush_mm_cond(mm)
+#define flush_tlb_kernel_range(start, end)	__tlb_flush_mm(&init_mm)
 
 #endif /* _S390_TLBFLUSH_H */
author	Linus Torvalds <torvalds@woody.linux-foundation.org>	2007-10-22 19:23:34 -0700
committer	Linus Torvalds <torvalds@woody.linux-foundation.org>	2007-10-22 19:23:34 -0700
commit	56d61a0e26c5a61c66d1ac259a59960295939da9 (patch)
tree	a23a30a966fe4220060682179294087cba1f9c57
parent	5f48b338cd28f4095697a174d7e3e72084aca893 (diff)
parent	190a1d722a59725706daf832bc8a511ed62f249d (diff)