8 files changed, 455 insertions, 122 deletions

diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
index 373679b3744a..d48ede334434 100644
--- a/arch/s390/Kconfig
+++ b/arch/s390/Kconfig
@@ -92,6 +92,9 @@ config S390
         select HAVE_ARCH_JUMP_LABEL if !MARCH_G5
         select HAVE_RCU_TABLE_FREE if SMP
         select ARCH_SAVE_PAGE_KEYS if HIBERNATION
+        select HAVE_MEMBLOCK
+        select HAVE_MEMBLOCK_NODE_MAP
+        select ARCH_DISCARD_MEMBLOCK
         select ARCH_INLINE_SPIN_TRYLOCK
         select ARCH_INLINE_SPIN_TRYLOCK_BH
         select ARCH_INLINE_SPIN_LOCK
@@ -345,9 +348,6 @@ config WARN_DYNAMIC_STACK
 
           Say N if you are unsure.
 
-config ARCH_POPULATES_NODE_MAP
-        def_bool y
-
 comment "Kernel preemption"
 
 source "kernel/Kconfig.preempt"
diff --git a/arch/s390/appldata/appldata_os.c b/arch/s390/appldata/appldata_os.c
index 92f1cb745d69..4de031d6b76c 100644
--- a/arch/s390/appldata/appldata_os.c
+++ b/arch/s390/appldata/appldata_os.c
@@ -115,21 +115,21 @@ static void appldata_get_os_data(void *data)
         j = 0;
         for_each_online_cpu(i) {
                 os_data->os_cpu[j].per_cpu_user =
-                        cputime_to_jiffies(kstat_cpu(i).cpustat.user);
+                        cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_USER]);
                 os_data->os_cpu[j].per_cpu_nice =
-                        cputime_to_jiffies(kstat_cpu(i).cpustat.nice);
+                        cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_NICE]);
                 os_data->os_cpu[j].per_cpu_system =
-                        cputime_to_jiffies(kstat_cpu(i).cpustat.system);
+                        cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_SYSTEM]);
                 os_data->os_cpu[j].per_cpu_idle =
-                        cputime_to_jiffies(kstat_cpu(i).cpustat.idle);
+                        cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_IDLE]);
                 os_data->os_cpu[j].per_cpu_irq =
-                        cputime_to_jiffies(kstat_cpu(i).cpustat.irq);
+                        cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_IRQ]);
                 os_data->os_cpu[j].per_cpu_softirq =
-                        cputime_to_jiffies(kstat_cpu(i).cpustat.softirq);
+                        cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_SOFTIRQ]);
                 os_data->os_cpu[j].per_cpu_iowait =
-                        cputime_to_jiffies(kstat_cpu(i).cpustat.iowait);
+                        cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_IOWAIT]);
                 os_data->os_cpu[j].per_cpu_steal =
-                        cputime_to_jiffies(kstat_cpu(i).cpustat.steal);
+                        cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_STEAL]);
                 os_data->os_cpu[j].cpu_id = i;
                 j++;
         }
diff --git a/arch/s390/include/asm/cputime.h b/arch/s390/include/asm/cputime.h
index 081434878296..c23c3900c304 100644
--- a/arch/s390/include/asm/cputime.h
+++ b/arch/s390/include/asm/cputime.h
@@ -16,114 +16,100 @@
 
 /* We want to use full resolution of the CPU timer: 2**-12 micro-seconds. */
 
-typedef unsigned long long cputime_t;
-typedef unsigned long long cputime64_t;
+typedef unsigned long long __nocast cputime_t;
+typedef unsigned long long __nocast cputime64_t;
 
-#ifndef __s390x__
-
-static inline unsigned int
-__div(unsigned long long n, unsigned int base)
+static inline unsigned long __div(unsigned long long n, unsigned long base)
 {
+#ifndef __s390x__
         register_pair rp;
 
         rp.pair = n >> 1;
         asm ("dr %0,%1" : "+d" (rp) : "d" (base >> 1));
         return rp.subreg.odd;
+#else /* __s390x__ */
+        return n / base;
+#endif /* __s390x__ */
 }
 
-#else /* __s390x__ */
+#define cputime_one_jiffy               jiffies_to_cputime(1)
+
+/*
+ * Convert cputime to jiffies and back.
+ */
+static inline unsigned long cputime_to_jiffies(const cputime_t cputime)
+{
+        return __div((__force unsigned long long) cputime, 4096000000ULL / HZ);
+}
 
-static inline unsigned int
-__div(unsigned long long n, unsigned int base)
+static inline cputime_t jiffies_to_cputime(const unsigned int jif)
 {
-        return n / base;
+        return (__force cputime_t)(jif * (4096000000ULL / HZ));
 }
 
-#endif /* __s390x__ */
+static inline u64 cputime64_to_jiffies64(cputime64_t cputime)
+{
+        unsigned long long jif = (__force unsigned long long) cputime;
+        do_div(jif, 4096000000ULL / HZ);
+        return jif;
+}
 
-#define cputime_zero                    (0ULL)
-#define cputime_one_jiffy               jiffies_to_cputime(1)
-#define cputime_max                     ((~0UL >> 1) - 1)
-#define cputime_add(__a, __b)           ((__a) +  (__b))
-#define cputime_sub(__a, __b)           ((__a) -  (__b))
-#define cputime_div(__a, __n) ({                \
-        unsigned long long __div = (__a);       \
-        do_div(__div,__n);                      \
-        __div;                                  \
-})
-#define cputime_halve(__a)              ((__a) >> 1)
-#define cputime_eq(__a, __b)            ((__a) == (__b))
-#define cputime_gt(__a, __b)            ((__a) >  (__b))
-#define cputime_ge(__a, __b)            ((__a) >= (__b))
-#define cputime_lt(__a, __b)            ((__a) <  (__b))
-#define cputime_le(__a, __b)            ((__a) <= (__b))
-#define cputime_to_jiffies(__ct)        (__div((__ct), 4096000000ULL / HZ))
-#define cputime_to_scaled(__ct)         (__ct)
-#define jiffies_to_cputime(__hz)        ((cputime_t)(__hz) * (4096000000ULL / HZ))
-
-#define cputime64_zero                  (0ULL)
-#define cputime64_add(__a, __b)         ((__a) + (__b))
-#define cputime_to_cputime64(__ct)      (__ct)
-
-static inline u64
-cputime64_to_jiffies64(cputime64_t cputime)
-{
-        do_div(cputime, 4096000000ULL / HZ);
-        return cputime;
+static inline cputime64_t jiffies64_to_cputime64(const u64 jif)
+{
+        return (__force cputime64_t)(jif * (4096000000ULL / HZ));
 }
 
 /*
  * Convert cputime to microseconds and back.
  */
-static inline unsigned int
-cputime_to_usecs(const cputime_t cputime)
+static inline unsigned int cputime_to_usecs(const cputime_t cputime)
 {
-        return cputime_div(cputime, 4096);
+        return (__force unsigned long long) cputime >> 12;
 }
 
-static inline cputime_t
-usecs_to_cputime(const unsigned int m)
+static inline cputime_t usecs_to_cputime(const unsigned int m)
 {
-        return (cputime_t) m * 4096;
+        return (__force cputime_t)(m * 4096ULL);
 }
 
+#define usecs_to_cputime64(m)           usecs_to_cputime(m)
+
 /*
  * Convert cputime to milliseconds and back.
  */
-static inline unsigned int
-cputime_to_secs(const cputime_t cputime)
+static inline unsigned int cputime_to_secs(const cputime_t cputime)
 {
-        return __div(cputime, 2048000000) >> 1;
+        return __div((__force unsigned long long) cputime, 2048000000) >> 1;
 }
 
-static inline cputime_t
-secs_to_cputime(const unsigned int s)
+static inline cputime_t secs_to_cputime(const unsigned int s)
 {
-        return (cputime_t) s * 4096000000ULL;
+        return (__force cputime_t)(s * 4096000000ULL);
 }
 
 /*
  * Convert cputime to timespec and back.
  */
-static inline cputime_t
-timespec_to_cputime(const struct timespec *value)
+static inline cputime_t timespec_to_cputime(const struct timespec *value)
 {
-        return value->tv_nsec * 4096 / 1000 + (u64) value->tv_sec * 4096000000ULL;
+        unsigned long long ret = value->tv_sec * 4096000000ULL;
+        return (__force cputime_t)(ret + value->tv_nsec * 4096 / 1000);
 }
 
-static inline void
-cputime_to_timespec(const cputime_t cputime, struct timespec *value)
+static inline void cputime_to_timespec(const cputime_t cputime,
+                                       struct timespec *value)
 {
+        unsigned long long __cputime = (__force unsigned long long) cputime;
 #ifndef __s390x__
         register_pair rp;
 
-        rp.pair = cputime >> 1;
+        rp.pair = __cputime >> 1;
         asm ("dr %0,%1" : "+d" (rp) : "d" (2048000000UL));
         value->tv_nsec = rp.subreg.even * 1000 / 4096;
         value->tv_sec = rp.subreg.odd;
 #else
-        value->tv_nsec = (cputime % 4096000000ULL) * 1000 / 4096;
-        value->tv_sec = cputime / 4096000000ULL;
+        value->tv_nsec = (__cputime % 4096000000ULL) * 1000 / 4096;
+        value->tv_sec = __cputime / 4096000000ULL;
 #endif
 }
 
@@ -132,50 +118,52 @@ cputime_to_timespec(const cputime_t cputime, struct timespec *value)
  * Since cputime and timeval have the same resolution (microseconds)
  * this is easy.
  */
-static inline cputime_t
-timeval_to_cputime(const struct timeval *value)
+static inline cputime_t timeval_to_cputime(const struct timeval *value)
 {
-        return value->tv_usec * 4096 + (u64) value->tv_sec * 4096000000ULL;
+        unsigned long long ret = value->tv_sec * 4096000000ULL;
+        return (__force cputime_t)(ret + value->tv_usec * 4096ULL);
 }
 
-static inline void
-cputime_to_timeval(const cputime_t cputime, struct timeval *value)
+static inline void cputime_to_timeval(const cputime_t cputime,
+                                      struct timeval *value)
 {
+        unsigned long long __cputime = (__force unsigned long long) cputime;
 #ifndef __s390x__
         register_pair rp;
 
-        rp.pair = cputime >> 1;
+        rp.pair = __cputime >> 1;
         asm ("dr %0,%1" : "+d" (rp) : "d" (2048000000UL));
         value->tv_usec = rp.subreg.even / 4096;
         value->tv_sec = rp.subreg.odd;
 #else
-        value->tv_usec = (cputime % 4096000000ULL) / 4096;
-        value->tv_sec = cputime / 4096000000ULL;
+        value->tv_usec = (__cputime % 4096000000ULL) / 4096;
+        value->tv_sec = __cputime / 4096000000ULL;
 #endif
 }
 
 /*
  * Convert cputime to clock and back.
  */
-static inline clock_t
-cputime_to_clock_t(cputime_t cputime)
+static inline clock_t cputime_to_clock_t(cputime_t cputime)
 {
-        return cputime_div(cputime, 4096000000ULL / USER_HZ);
+        unsigned long long clock = (__force unsigned long long) cputime;
+        do_div(clock, 4096000000ULL / USER_HZ);
+        return clock;
 }
 
-static inline cputime_t
-clock_t_to_cputime(unsigned long x)
+static inline cputime_t clock_t_to_cputime(unsigned long x)
 {
-        return (cputime_t) x * (4096000000ULL / USER_HZ);
+        return (__force cputime_t)(x * (4096000000ULL / USER_HZ));
 }
 
 /*
  * Convert cputime64 to clock.
  */
-static inline clock_t
-cputime64_to_clock_t(cputime64_t cputime)
+static inline clock_t cputime64_to_clock_t(cputime64_t cputime)
 {
-       return cputime_div(cputime, 4096000000ULL / USER_HZ);
+        unsigned long long clock = (__force unsigned long long) cputime;
+        do_div(clock, 4096000000ULL / USER_HZ);
+        return clock;
 }
 
 struct s390_idle_data {
diff --git a/arch/s390/kernel/process.c b/arch/s390/kernel/process.c
index 9451b210a1b4..3201ae447990 100644
--- a/arch/s390/kernel/process.c
+++ b/arch/s390/kernel/process.c
@@ -91,10 +91,12 @@ static void default_idle(void)
 void cpu_idle(void)
 {
         for (;;) {
-                tick_nohz_stop_sched_tick(1);
+                tick_nohz_idle_enter();
+                rcu_idle_enter();
                 while (!need_resched())
                         default_idle();
-                tick_nohz_restart_sched_tick();
+                rcu_idle_exit();
+                tick_nohz_idle_exit();
                 preempt_enable_no_resched();
                 schedule();
                 preempt_disable();
diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c
index e54c4ff8abaa..f11d1b037c50 100644
--- a/arch/s390/kernel/setup.c
+++ b/arch/s390/kernel/setup.c
@@ -21,6 +21,7 @@
 #include <linux/module.h>
 #include <linux/sched.h>
 #include <linux/kernel.h>
+#include <linux/memblock.h>
 #include <linux/mm.h>
 #include <linux/stddef.h>
 #include <linux/unistd.h>
@@ -820,7 +821,8 @@ setup_memory(void)
                 end_chunk = min(end_chunk, end_pfn);
                 if (start_chunk >= end_chunk)
                         continue;
-                add_active_range(0, start_chunk, end_chunk);
+                memblock_add_node(PFN_PHYS(start_chunk),
+                                  PFN_PHYS(end_chunk - start_chunk), 0);
                 pfn = max(start_chunk, start_pfn);
                 for (; pfn < end_chunk; pfn++)
                         page_set_storage_key(PFN_PHYS(pfn),
diff --git a/arch/s390/oprofile/hwsampler.c b/arch/s390/oprofile/hwsampler.c
index f43c0e4282af..9daee91e6c3f 100644
--- a/arch/s390/oprofile/hwsampler.c
+++ b/arch/s390/oprofile/hwsampler.c
@@ -22,6 +22,7 @@
 #include <asm/irq.h>
 
 #include "hwsampler.h"
+#include "op_counter.h"
 
 #define MAX_NUM_SDB 511
 #define MIN_NUM_SDB 1
@@ -896,6 +897,8 @@ static void add_samples_to_oprofile(unsigned int cpu, unsigned long *sdbt,
                 if (sample_data_ptr->P == 1) {
                         /* userspace sample */
                         unsigned int pid = sample_data_ptr->prim_asn;
+                        if (!counter_config.user)
+                                goto skip_sample;
                         rcu_read_lock();
                         tsk = pid_task(find_vpid(pid), PIDTYPE_PID);
                         if (tsk)
@@ -903,6 +906,8 @@ static void add_samples_to_oprofile(unsigned int cpu, unsigned long *sdbt,
                         rcu_read_unlock();
                 } else {
                         /* kernelspace sample */
+                        if (!counter_config.kernel)
+                                goto skip_sample;
                         regs = task_pt_regs(current);
                 }
 
@@ -910,7 +915,7 @@ static void add_samples_to_oprofile(unsigned int cpu, unsigned long *sdbt,
                 oprofile_add_ext_hw_sample(sample_data_ptr->ia, regs, 0,
                                 !sample_data_ptr->P, tsk);
                 mutex_unlock(&hws_sem);
-
+        skip_sample:
                 sample_data_ptr++;
         }
 }
diff --git a/arch/s390/oprofile/init.c b/arch/s390/oprofile/init.c
index bd58b72454cf..2297be406c61 100644
--- a/arch/s390/oprofile/init.c
+++ b/arch/s390/oprofile/init.c
@@ -2,10 +2,11 @@
  * arch/s390/oprofile/init.c
  *
  * S390 Version
- *   Copyright (C) 2003 IBM Deutschland Entwicklung GmbH, IBM Corporation
+ *   Copyright (C) 2002-2011 IBM Deutschland Entwicklung GmbH, IBM Corporation
  *   Author(s): Thomas Spatzier (tspat@de.ibm.com)
  *   Author(s): Mahesh Salgaonkar (mahesh@linux.vnet.ibm.com)
  *   Author(s): Heinz Graalfs (graalfs@linux.vnet.ibm.com)
+ *   Author(s): Andreas Krebbel (krebbel@linux.vnet.ibm.com)
  *
  * @remark Copyright 2002-2011 OProfile authors
  */
@@ -14,6 +15,8 @@
 #include <linux/init.h>
 #include <linux/errno.h>
 #include <linux/fs.h>
+#include <linux/module.h>
+#include <asm/processor.h>
 
 #include "../../../drivers/oprofile/oprof.h"
 
@@ -22,6 +25,7 @@ extern void s390_backtrace(struct pt_regs * const regs, unsigned int depth);
 #ifdef CONFIG_64BIT
 
 #include "hwsampler.h"
+#include "op_counter.h"
 
 #define DEFAULT_INTERVAL        4127518
 
@@ -35,16 +39,41 @@ static unsigned long oprofile_max_interval;
 static unsigned long oprofile_sdbt_blocks = DEFAULT_SDBT_BLOCKS;
 static unsigned long oprofile_sdb_blocks = DEFAULT_SDB_BLOCKS;
 
-static int hwsampler_file;
+static int hwsampler_enabled;
 static int hwsampler_running;   /* start_mutex must be held to change */
+static int hwsampler_available;
 
 static struct oprofile_operations timer_ops;
 
+struct op_counter_config counter_config;
+
+enum __force_cpu_type {
+        reserved = 0,           /* do not force */
+        timer,
+};
+static int force_cpu_type;
+
+static int set_cpu_type(const char *str, struct kernel_param *kp)
+{
+        if (!strcmp(str, "timer")) {
+                force_cpu_type = timer;
+                printk(KERN_INFO "oprofile: forcing timer to be returned "
+                                 "as cpu type\n");
+        } else {
+                force_cpu_type = 0;
+        }
+
+        return 0;
+}
+module_param_call(cpu_type, set_cpu_type, NULL, NULL, 0);
+MODULE_PARM_DESC(cpu_type, "Force legacy basic mode sampling"
+                           "(report cpu_type \"timer\"");
+
 static int oprofile_hwsampler_start(void)
 {
         int retval;
 
-        hwsampler_running = hwsampler_file;
+        hwsampler_running = hwsampler_enabled;
 
         if (!hwsampler_running)
                 return timer_ops.start();
@@ -72,10 +101,16 @@ static void oprofile_hwsampler_stop(void)
         return;
 }
 
+/*
+ * File ops used for:
+ * /dev/oprofile/0/enabled
+ * /dev/oprofile/hwsampling/hwsampler  (cpu_type = timer)
+ */
+
 static ssize_t hwsampler_read(struct file *file, char __user *buf,
                 size_t count, loff_t *offset)
 {
-        return oprofilefs_ulong_to_user(hwsampler_file, buf, count, offset);
+        return oprofilefs_ulong_to_user(hwsampler_enabled, buf, count, offset);
 }
 
 static ssize_t hwsampler_write(struct file *file, char const __user *buf,
@@ -91,6 +126,9 @@ static ssize_t hwsampler_write(struct file *file, char const __user *buf,
         if (retval <= 0)
                 return retval;
 
+        if (val != 0 && val != 1)
+                return -EINVAL;
+
         if (oprofile_started)
                 /*
                  * save to do without locking as we set
@@ -99,7 +137,7 @@ static ssize_t hwsampler_write(struct file *file, char const __user *buf,
                  */
                 return -EBUSY;
 
-        hwsampler_file = val;
+        hwsampler_enabled = val;
 
         return count;
 }
@@ -109,38 +147,311 @@ static const struct file_operations hwsampler_fops = {
         .write          = hwsampler_write,
 };
 
+/*
+ * File ops used for:
+ * /dev/oprofile/0/count
+ * /dev/oprofile/hwsampling/hw_interval  (cpu_type = timer)
+ *
+ * Make sure that the value is within the hardware range.
+ */
+
+static ssize_t hw_interval_read(struct file *file, char __user *buf,
+                                size_t count, loff_t *offset)
+{
+        return oprofilefs_ulong_to_user(oprofile_hw_interval, buf,
+                                        count, offset);
+}
+
+static ssize_t hw_interval_write(struct file *file, char const __user *buf,
+                                 size_t count, loff_t *offset)
+{
+        unsigned long val;
+        int retval;
+
+        if (*offset)
+                return -EINVAL;
+        retval = oprofilefs_ulong_from_user(&val, buf, count);
+        if (retval)
+                return retval;
+        if (val < oprofile_min_interval)
+                oprofile_hw_interval = oprofile_min_interval;
+        else if (val > oprofile_max_interval)
+                oprofile_hw_interval = oprofile_max_interval;
+        else
+                oprofile_hw_interval = val;
+
+        return count;
+}
+
+static const struct file_operations hw_interval_fops = {
+        .read           = hw_interval_read,
+        .write          = hw_interval_write,
+};
+
+/*
+ * File ops used for:
+ * /dev/oprofile/0/event
+ * Only a single event with number 0 is supported with this counter.
+ *
+ * /dev/oprofile/0/unit_mask
+ * This is a dummy file needed by the user space tools.
+ * No value other than 0 is accepted or returned.
+ */
+
+static ssize_t hwsampler_zero_read(struct file *file, char __user *buf,
+                                    size_t count, loff_t *offset)
+{
+        return oprofilefs_ulong_to_user(0, buf, count, offset);
+}
+
+static ssize_t hwsampler_zero_write(struct file *file, char const __user *buf,
+                                     size_t count, loff_t *offset)
+{
+        unsigned long val;
+        int retval;
+
+        if (*offset)
+                return -EINVAL;
+
+        retval = oprofilefs_ulong_from_user(&val, buf, count);
+        if (retval)
+                return retval;
+        if (val != 0)
+                return -EINVAL;
+        return count;
+}
+
+static const struct file_operations zero_fops = {
+        .read           = hwsampler_zero_read,
+        .write          = hwsampler_zero_write,
+};
+
+/* /dev/oprofile/0/kernel file ops.  */
+
+static ssize_t hwsampler_kernel_read(struct file *file, char __user *buf,
+                                     size_t count, loff_t *offset)
+{
+        return oprofilefs_ulong_to_user(counter_config.kernel,
+                                        buf, count, offset);
+}
+
+static ssize_t hwsampler_kernel_write(struct file *file, char const __user *buf,
+                                      size_t count, loff_t *offset)
+{
+        unsigned long val;
+        int retval;
+
+        if (*offset)
+                return -EINVAL;
+
+        retval = oprofilefs_ulong_from_user(&val, buf, count);
+        if (retval)
+                return retval;
+
+        if (val != 0 && val != 1)
+                return -EINVAL;
+
+        counter_config.kernel = val;
+
+        return count;
+}
+
+static const struct file_operations kernel_fops = {
+        .read           = hwsampler_kernel_read,
+        .write          = hwsampler_kernel_write,
+};
+
+/* /dev/oprofile/0/user file ops. */
+
+static ssize_t hwsampler_user_read(struct file *file, char __user *buf,
+                                   size_t count, loff_t *offset)
+{
+        return oprofilefs_ulong_to_user(counter_config.user,
+                                        buf, count, offset);
+}
+
+static ssize_t hwsampler_user_write(struct file *file, char const __user *buf,
+                                    size_t count, loff_t *offset)
+{
+        unsigned long val;
+        int retval;
+
+        if (*offset)
+                return -EINVAL;
+
+        retval = oprofilefs_ulong_from_user(&val, buf, count);
+        if (retval)
+                return retval;
+
+        if (val != 0 && val != 1)
+                return -EINVAL;
+
+        counter_config.user = val;
+
+        return count;
+}
+
+static const struct file_operations user_fops = {
+        .read           = hwsampler_user_read,
+        .write          = hwsampler_user_write,
+};
+
+
+/*
+ * File ops used for: /dev/oprofile/timer/enabled
+ * The value always has to be the inverted value of hwsampler_enabled. So
+ * no separate variable is created. That way we do not need locking.
+ */
+
+static ssize_t timer_enabled_read(struct file *file, char __user *buf,
+                                  size_t count, loff_t *offset)
+{
+        return oprofilefs_ulong_to_user(!hwsampler_enabled, buf, count, offset);
+}
+
+static ssize_t timer_enabled_write(struct file *file, char const __user *buf,
+                                   size_t count, loff_t *offset)
+{
+        unsigned long val;
+        int retval;
+
+        if (*offset)
+                return -EINVAL;
+
+        retval = oprofilefs_ulong_from_user(&val, buf, count);
+        if (retval)
+                return retval;
+
+        if (val != 0 && val != 1)
+                return -EINVAL;
+
+        /* Timer cannot be disabled without having hardware sampling.  */
+        if (val == 0 && !hwsampler_available)
+                return -EINVAL;
+
+        if (oprofile_started)
+                /*
+                 * save to do without locking as we set
+                 * hwsampler_running in start() when start_mutex is
+                 * held
+                 */
+                return -EBUSY;
+
+        hwsampler_enabled = !val;
+
+        return count;
+}
+
+static const struct file_operations timer_enabled_fops = {
+        .read           = timer_enabled_read,
+        .write          = timer_enabled_write,
+};
+
+
 static int oprofile_create_hwsampling_files(struct super_block *sb,
-                                                struct dentry *root)
+                                            struct dentry *root)
 {
-        struct dentry *hw_dir;
+        struct dentry *dir;
+
+        dir = oprofilefs_mkdir(sb, root, "timer");
+        if (!dir)
+                return -EINVAL;
+
+        oprofilefs_create_file(sb, dir, "enabled", &timer_enabled_fops);
+
+        if (!hwsampler_available)
+                return 0;
 
         /* reinitialize default values */
-        hwsampler_file = 1;
+        hwsampler_enabled = 1;
+        counter_config.kernel = 1;
+        counter_config.user = 1;
 
-        hw_dir = oprofilefs_mkdir(sb, root, "hwsampling");
-        if (!hw_dir)
-                return -EINVAL;
+        if (!force_cpu_type) {
+                /*
+                 * Create the counter file system.  A single virtual
+                 * counter is created which can be used to
+                 * enable/disable hardware sampling dynamically from
+                 * user space.  The user space will configure a single
+                 * counter with a single event.  The value of 'event'
+                 * and 'unit_mask' are not evaluated by the kernel code
+                 * and can only be set to 0.
+                 */
+
+                dir = oprofilefs_mkdir(sb, root, "0");
+                if (!dir)
+                        return -EINVAL;
 
-        oprofilefs_create_file(sb, hw_dir, "hwsampler", &hwsampler_fops);
-        oprofilefs_create_ulong(sb, hw_dir, "hw_interval",
-                                &oprofile_hw_interval);
-        oprofilefs_create_ro_ulong(sb, hw_dir, "hw_min_interval",
-                                &oprofile_min_interval);
-        oprofilefs_create_ro_ulong(sb, hw_dir, "hw_max_interval",
-                                &oprofile_max_interval);
-        oprofilefs_create_ulong(sb, hw_dir, "hw_sdbt_blocks",
-                                &oprofile_sdbt_blocks);
+                oprofilefs_create_file(sb, dir, "enabled", &hwsampler_fops);
+                oprofilefs_create_file(sb, dir, "event", &zero_fops);
+                oprofilefs_create_file(sb, dir, "count", &hw_interval_fops);
+                oprofilefs_create_file(sb, dir, "unit_mask", &zero_fops);
+                oprofilefs_create_file(sb, dir, "kernel", &kernel_fops);
+                oprofilefs_create_file(sb, dir, "user", &user_fops);
+                oprofilefs_create_ulong(sb, dir, "hw_sdbt_blocks",
+                                        &oprofile_sdbt_blocks);
 
+        } else {
+                /*
+                 * Hardware sampling can be used but the cpu_type is
+                 * forced to timer in order to deal with legacy user
+                 * space tools.  The /dev/oprofile/hwsampling fs is
+                 * provided in that case.
+                 */
+                dir = oprofilefs_mkdir(sb, root, "hwsampling");
+                if (!dir)
+                        return -EINVAL;
+
+                oprofilefs_create_file(sb, dir, "hwsampler",
+                                       &hwsampler_fops);
+                oprofilefs_create_file(sb, dir, "hw_interval",
+                                       &hw_interval_fops);
+                oprofilefs_create_ro_ulong(sb, dir, "hw_min_interval",
+                                           &oprofile_min_interval);
+                oprofilefs_create_ro_ulong(sb, dir, "hw_max_interval",
+                                           &oprofile_max_interval);
+                oprofilefs_create_ulong(sb, dir, "hw_sdbt_blocks",
+                                        &oprofile_sdbt_blocks);
+        }
         return 0;
 }
 
 static int oprofile_hwsampler_init(struct oprofile_operations *ops)
 {
+        /*
+         * Initialize the timer mode infrastructure as well in order
+         * to be able to switch back dynamically.  oprofile_timer_init
+         * is not supposed to fail.
+         */
+        if (oprofile_timer_init(ops))
+                BUG();
+
+        memcpy(&timer_ops, ops, sizeof(timer_ops));
+        ops->create_files = oprofile_create_hwsampling_files;
+
+        /*
+         * If the user space tools do not support newer cpu types,
+         * the force_cpu_type module parameter
+         * can be used to always return \"timer\" as cpu type.
+         */
+        if (force_cpu_type != timer) {
+                struct cpuid id;
+
+                get_cpu_id (&id);
+
+                switch (id.machine) {
+                case 0x2097: case 0x2098: ops->cpu_type = "s390/z10"; break;
+                case 0x2817: case 0x2818: ops->cpu_type = "s390/z196"; break;
+                default: return -ENODEV;
+                }
+        }
+
         if (hwsampler_setup())
                 return -ENODEV;
 
         /*
-         * create hwsampler files only if hwsampler_setup() succeeds.
+         * Query the range for the sampling interval from the
+         * hardware.
          */
         oprofile_min_interval = hwsampler_query_min_interval();
         if (oprofile_min_interval == 0)
@@ -155,23 +466,17 @@ static int oprofile_hwsampler_init(struct oprofile_operations *ops)
         if (oprofile_hw_interval > oprofile_max_interval)
                 oprofile_hw_interval = oprofile_max_interval;
 
-        if (oprofile_timer_init(ops))
-                return -ENODEV;
-
-        printk(KERN_INFO "oprofile: using hardware sampling\n");
-
-        memcpy(&timer_ops, ops, sizeof(timer_ops));
+        printk(KERN_INFO "oprofile: System z hardware sampling "
+               "facility found.\n");
 
         ops->start = oprofile_hwsampler_start;
         ops->stop = oprofile_hwsampler_stop;
-        ops->create_files = oprofile_create_hwsampling_files;
 
         return 0;
 }
 
 static void oprofile_hwsampler_exit(void)
 {
-        oprofile_timer_exit();
         hwsampler_shutdown();
 }
 
@@ -182,7 +487,15 @@ int __init oprofile_arch_init(struct oprofile_operations *ops)
         ops->backtrace = s390_backtrace;
 
 #ifdef CONFIG_64BIT
-        return oprofile_hwsampler_init(ops);
+
+        /*
+         * -ENODEV is not reported to the caller.  The module itself
+         * will use the timer mode sampling as fallback and this is
+         * always available.
+         */
+        hwsampler_available = oprofile_hwsampler_init(ops) == 0;
+
+        return 0;
 #else
         return -ENODEV;
 #endif
diff --git a/arch/s390/oprofile/op_counter.h b/arch/s390/oprofile/op_counter.h
new file mode 100644
index 000000000000..1a8d3ca09014
--- /dev/null
+++ b/arch/s390/oprofile/op_counter.h
@@ -0,0 +1,23 @@
+/**
+ * arch/s390/oprofile/op_counter.h
+ *
+ *   Copyright (C) 2011 IBM Deutschland Entwicklung GmbH, IBM Corporation
+ *   Author(s): Andreas Krebbel (krebbel@linux.vnet.ibm.com)
+ *
+ * @remark Copyright 2011 OProfile authors
+ */
+
+#ifndef OP_COUNTER_H
+#define OP_COUNTER_H
+
+struct op_counter_config {
+        /* `enabled' maps to the hwsampler_file variable.  */
+        /* `count' maps to the oprofile_hw_interval variable.  */
+        /* `event' and `unit_mask' are unused. */
+        unsigned long kernel;
+        unsigned long user;
+};
+
+extern struct op_counter_config counter_config;
+
+#endif /* OP_COUNTER_H */