5 files changed, 440 insertions, 206 deletions

diff --git a/arch/s390/oprofile/backtrace.c b/arch/s390/oprofile/backtrace.c
index bc4b84a35ca..8a6811b2cdb 100644
--- a/arch/s390/oprofile/backtrace.c
+++ b/arch/s390/oprofile/backtrace.c
@@ -1,8 +1,6 @@
-/**
- * arch/s390/oprofile/backtrace.c
- *
+/*
  * S390 Version
- *   Copyright (C) 2005 IBM Corporation, IBM Deutschland Entwicklung GmbH.
+ *   Copyright IBM Corp. 2005
  *   Author(s): Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
  */
 
@@ -60,7 +58,7 @@ void s390_backtrace(struct pt_regs * const regs, unsigned int depth)
 	unsigned long head;
 	struct stack_frame* head_sf;
 
-	if (user_mode (regs))
+	if (user_mode(regs))
 		return;
 
 	head = regs->gprs[15];
diff --git a/arch/s390/oprofile/hwsampler.c b/arch/s390/oprofile/hwsampler.c
index 4552ce40c81..e53c6f26880 100644
--- a/arch/s390/oprofile/hwsampler.c
+++ b/arch/s390/oprofile/hwsampler.c
@@ -1,6 +1,4 @@
-/**
- * arch/s390/oprofile/hwsampler.c
- *
+/*
  * Copyright IBM Corp. 2010
  * Author: Heinz Graalfs <graalfs@de.ibm.com>
  */
@@ -18,23 +16,16 @@
 #include <linux/oom.h>
 #include <linux/oprofile.h>
 
-#include <asm/lowcore.h>
+#include <asm/facility.h>
+#include <asm/cpu_mf.h>
 #include <asm/irq.h>
 
 #include "hwsampler.h"
+#include "op_counter.h"
 
 #define MAX_NUM_SDB 511
 #define MIN_NUM_SDB 1
 
-#define ALERT_REQ_MASK   0x4000000000000000ul
-#define BUFFER_FULL_MASK 0x8000000000000000ul
-
-#define EI_IEA      (1 << 31)	/* invalid entry address              */
-#define EI_ISE      (1 << 30)	/* incorrect SDBT entry               */
-#define EI_PRA      (1 << 29)	/* program request alert              */
-#define EI_SACA     (1 << 23)	/* sampler authorization change alert */
-#define EI_LSDA     (1 << 22)	/* loss of sample data alert          */
-
 DECLARE_PER_CPU(struct hws_cpu_buffer, sampler_cpu_buffer);
 
 struct hws_execute_parms {
@@ -50,6 +41,7 @@ static DEFINE_MUTEX(hws_sem_oom);
 
 static unsigned char hws_flush_all;
 static unsigned int hws_oom;
+static unsigned int hws_alert;
 static struct workqueue_struct *hws_wq;
 
 static unsigned int hws_state;
@@ -71,43 +63,6 @@ static unsigned long interval;
 static unsigned long min_sampler_rate;
 static unsigned long max_sampler_rate;
 
-static int ssctl(void *buffer)
-{
-	int cc;
-
-	/* set in order to detect a program check */
-	cc = 1;
-
-	asm volatile(
-		"0: .insn s,0xB2870000,0(%1)\n"
-		"1: ipm %0\n"
-		"   srl %0,28\n"
-		"2:\n"
-		EX_TABLE(0b, 2b) EX_TABLE(1b, 2b)
-		: "+d" (cc), "+a" (buffer)
-		: "m" (*((struct hws_ssctl_request_block *)buffer))
-		: "cc", "memory");
-
-	return cc ? -EINVAL : 0 ;
-}
-
-static int qsi(void *buffer)
-{
-	int cc;
-	cc = 1;
-
-	asm volatile(
-		"0: .insn s,0xB2860000,0(%1)\n"
-		"1: lhi %0,0\n"
-		"2:\n"
-		EX_TABLE(0b, 2b) EX_TABLE(1b, 2b)
-		: "=d" (cc), "+a" (buffer)
-		: "m" (*((struct hws_qsi_info_block *)buffer))
-		: "cc", "memory");
-
-	return cc ? -EINVAL : 0;
-}
-
 static void execute_qsi(void *parms)
 {
 	struct hws_execute_parms *ep = parms;
@@ -119,7 +74,7 @@ static void execute_ssctl(void *parms)
 {
 	struct hws_execute_parms *ep = parms;
 
-	ep->rc = ssctl(ep->buffer);
+	ep->rc = lsctl(ep->buffer);
 }
 
 static int smp_ctl_ssctl_stop(int cpu)
@@ -220,20 +175,23 @@ static int smp_ctl_qsi(int cpu)
 	return ep.rc;
 }
 
-static inline unsigned long *trailer_entry_ptr(unsigned long v)
+static void hws_ext_handler(struct ext_code ext_code,
+			    unsigned int param32, unsigned long param64)
 {
-	void *ret;
+	struct hws_cpu_buffer *cb = &__get_cpu_var(sampler_cpu_buffer);
 
-	ret = (void *)v;
-	ret += PAGE_SIZE;
-	ret -= sizeof(struct hws_trailer_entry);
+	if (!(param32 & CPU_MF_INT_SF_MASK))
+		return;
 
-	return (unsigned long *) ret;
-}
+	if (!hws_alert)
+		return;
 
-/* prototypes for external interrupt handler and worker */
-static void hws_ext_handler(unsigned int ext_int_code,
-				unsigned int param32, unsigned long param64);
+	inc_irq_stat(IRQEXT_CMS);
+	atomic_xchg(&cb->ext_params, atomic_read(&cb->ext_params) | param32);
+
+	if (hws_wq)
+		queue_work(hws_wq, &cb->worker);
+}
 
 static void worker(struct work_struct *work);
 
@@ -251,23 +209,11 @@ static void init_all_cpu_buffers(void)
 	}
 }
 
-static int is_link_entry(unsigned long *s)
-{
-	return *s & 0x1ul ? 1 : 0;
-}
-
-static unsigned long *get_next_sdbt(unsigned long *s)
+static void prepare_cpu_buffers(void)
 {
-	return (unsigned long *) (*s & ~0x1ul);
-}
-
-static int prepare_cpu_buffers(void)
-{
-	int cpu;
-	int rc;
 	struct hws_cpu_buffer *cb;
+	int cpu;
 
-	rc = 0;
 	for_each_online_cpu(cpu) {
 		cb = &per_cpu(sampler_cpu_buffer, cpu);
 		atomic_set(&cb->ext_params, 0);
@@ -282,8 +228,6 @@ static int prepare_cpu_buffers(void)
 		cb->oom = 0;
 		cb->stop_mode = 0;
 	}
-
-	return rc;
 }
 
 /*
@@ -348,7 +292,7 @@ static int allocate_sdbt(int cpu)
 			}
 			*sdbt = sdb;
 			trailer = trailer_entry_ptr(*sdbt);
-			*trailer = ALERT_REQ_MASK;
+			*trailer = SDB_TE_ALERT_REQ_MASK;
 			sdbt++;
 			mutex_unlock(&hws_sem_oom);
 		}
@@ -672,18 +616,6 @@ int hwsampler_activate(unsigned int cpu)
 	return rc;
 }
 
-static void hws_ext_handler(unsigned int ext_int_code,
-			    unsigned int param32, unsigned long param64)
-{
-	struct hws_cpu_buffer *cb;
-
-	kstat_cpu(smp_processor_id()).irqs[EXTINT_CPM]++;
-	cb = &__get_cpu_var(sampler_cpu_buffer);
-	atomic_xchg(&cb->ext_params, atomic_read(&cb->ext_params) | param32);
-	if (hws_wq)
-		queue_work(hws_wq, &cb->worker);
-}
-
 static int check_qsi_on_setup(void)
 {
 	int rc;
@@ -759,23 +691,23 @@ static int worker_check_error(unsigned int cpu, int ext_params)
 	if (!sdbt || !*sdbt)
 		return -EINVAL;
 
-	if (ext_params & EI_PRA)
+	if (ext_params & CPU_MF_INT_SF_PRA)
 		cb->req_alert++;
 
-	if (ext_params & EI_LSDA)
+	if (ext_params & CPU_MF_INT_SF_LSDA)
 		cb->loss_of_sample_data++;
 
-	if (ext_params & EI_IEA) {
+	if (ext_params & CPU_MF_INT_SF_IAE) {
 		cb->invalid_entry_address++;
 		rc = -EINVAL;
 	}
 
-	if (ext_params & EI_ISE) {
+	if (ext_params & CPU_MF_INT_SF_ISE) {
 		cb->incorrect_sdbt_entry++;
 		rc = -EINVAL;
 	}
 
-	if (ext_params & EI_SACA) {
+	if (ext_params & CPU_MF_INT_SF_SACA) {
 		cb->sample_auth_change_alert++;
 		rc = -EINVAL;
 	}
@@ -836,7 +768,7 @@ static void worker_on_interrupt(unsigned int cpu)
 
 		trailer = trailer_entry_ptr(*sdbt);
 		/* leave loop if no more work to do */
-		if (!(*trailer & BUFFER_FULL_MASK)) {
+		if (!(*trailer & SDB_TE_BUFFER_FULL_MASK)) {
 			done = 1;
 			if (!hws_flush_all)
 				continue;
@@ -863,7 +795,7 @@ static void worker_on_interrupt(unsigned int cpu)
 static void add_samples_to_oprofile(unsigned int cpu, unsigned long *sdbt,
 		unsigned long *dear)
 {
-	struct hws_data_entry *sample_data_ptr;
+	struct hws_basic_entry *sample_data_ptr;
 	unsigned long *trailer;
 
 	trailer = trailer_entry_ptr(*sdbt);
@@ -873,7 +805,7 @@ static void add_samples_to_oprofile(unsigned int cpu, unsigned long *sdbt,
 		trailer = dear;
 	}
 
-	sample_data_ptr = (struct hws_data_entry *)(*sdbt);
+	sample_data_ptr = (struct hws_basic_entry *)(*sdbt);
 
 	while ((unsigned long *)sample_data_ptr < trailer) {
 		struct pt_regs *regs = NULL;
@@ -896,6 +828,8 @@ 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 +837,8 @@ 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 +846,7 @@ 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++;
 	}
 }
@@ -994,7 +930,7 @@ allocate_error:
  *
  * Returns 0 on success, !0 on failure.
  */
-int hwsampler_deallocate()
+int hwsampler_deallocate(void)
 {
 	int rc;
 
@@ -1004,7 +940,8 @@ int hwsampler_deallocate()
 	if (hws_state != HWS_STOPPED)
 		goto deallocate_exit;
 
-	ctl_clear_bit(0, 5); /* set bit 58 CR0 off */
+	irq_subclass_unregister(IRQ_SUBCLASS_MEASUREMENT_ALERT);
+	hws_alert = 0;
 	deallocate_sdbt();
 
 	hws_state = HWS_DEALLOCATED;
@@ -1035,7 +972,7 @@ unsigned long hwsampler_get_sample_overflow_count(unsigned int cpu)
 	return cb->sample_overflow;
 }
 
-int hwsampler_setup()
+int hwsampler_setup(void)
 {
 	int rc;
 	int cpu;
@@ -1092,7 +1029,7 @@ int hwsampler_setup()
 				max_sampler_rate = cb->qsi.max_sampl_rate;
 		}
 	}
-	register_external_interrupt(0x1407, hws_ext_handler);
+	register_external_irq(EXT_IRQ_MEASURE_ALERT, hws_ext_handler);
 
 	hws_state = HWS_DEALLOCATED;
 	rc = 0;
@@ -1102,7 +1039,7 @@ setup_exit:
 	return rc;
 }
 
-int hwsampler_shutdown()
+int hwsampler_shutdown(void)
 {
 	int rc;
 
@@ -1118,7 +1055,8 @@ int hwsampler_shutdown()
 		mutex_lock(&hws_sem);
 
 		if (hws_state == HWS_STOPPED) {
-			ctl_clear_bit(0, 5); /* set bit 58 CR0 off */
+			irq_subclass_unregister(IRQ_SUBCLASS_MEASUREMENT_ALERT);
+			hws_alert = 0;
 			deallocate_sdbt();
 		}
 		if (hws_wq) {
@@ -1126,7 +1064,7 @@ int hwsampler_shutdown()
 			hws_wq = NULL;
 		}
 
-		unregister_external_interrupt(0x1407, hws_ext_handler);
+		unregister_external_irq(EXT_IRQ_MEASURE_ALERT, hws_ext_handler);
 		hws_state = HWS_INIT;
 		rc = 0;
 	}
@@ -1165,9 +1103,7 @@ int hwsampler_start_all(unsigned long rate)
 	if (rc)
 		goto start_all_exit;
 
-	rc = prepare_cpu_buffers();
-	if (rc)
-		goto start_all_exit;
+	prepare_cpu_buffers();
 
 	for_each_online_cpu(cpu) {
 		rc = start_sampling(cpu);
@@ -1193,7 +1129,8 @@ start_all_exit:
 	hws_oom = 1;
 	hws_flush_all = 0;
 	/* now let them in, 1407 CPUMF external interrupts */
-	ctl_set_bit(0, 5); /* set CR0 bit 58 */
+	hws_alert = 1;
+	irq_subclass_register(IRQ_SUBCLASS_MEASUREMENT_ALERT);
 
 	return 0;
 }
@@ -1203,7 +1140,7 @@ start_all_exit:
  *
  * Returns 0 on success, !0 on failure.
  */
-int hwsampler_stop_all()
+int hwsampler_stop_all(void)
 {
 	int tmp_rc, rc, cpu;
 	struct hws_cpu_buffer *cb;
@@ -1213,7 +1150,7 @@ int hwsampler_stop_all()
 	rc = 0;
 	if (hws_state == HWS_INIT) {
 		mutex_unlock(&hws_sem);
-		return rc;
+		return 0;
 	}
 	hws_state = HWS_STOPPING;
 	mutex_unlock(&hws_sem);
diff --git a/arch/s390/oprofile/hwsampler.h b/arch/s390/oprofile/hwsampler.h
index 1912f3bb190..a483d06f2fa 100644
--- a/arch/s390/oprofile/hwsampler.h
+++ b/arch/s390/oprofile/hwsampler.h
@@ -9,27 +9,7 @@
 #define HWSAMPLER_H_
 
 #include <linux/workqueue.h>
-
-struct hws_qsi_info_block          /* QUERY SAMPLING information block  */
-{ /* Bit(s) */
-	unsigned int b0_13:14;      /* 0-13: zeros                       */
-	unsigned int as:1;          /* 14: sampling authorisation control*/
-	unsigned int b15_21:7;      /* 15-21: zeros                      */
-	unsigned int es:1;          /* 22: sampling enable control       */
-	unsigned int b23_29:7;      /* 23-29: zeros                      */
-	unsigned int cs:1;          /* 30: sampling activation control   */
-	unsigned int:1;             /* 31: reserved                      */
-	unsigned int bsdes:16;      /* 4-5: size of sampling entry       */
-	unsigned int:16;            /* 6-7: reserved                     */
-	unsigned long min_sampl_rate; /* 8-15: minimum sampling interval */
-	unsigned long max_sampl_rate; /* 16-23: maximum sampling interval*/
-	unsigned long tear;         /* 24-31: TEAR contents              */
-	unsigned long dear;         /* 32-39: DEAR contents              */
-	unsigned int rsvrd0;        /* 40-43: reserved                   */
-	unsigned int cpu_speed;     /* 44-47: CPU speed                  */
-	unsigned long long rsvrd1;  /* 48-55: reserved                   */
-	unsigned long long rsvrd2;  /* 56-63: reserved                   */
-};
+#include <asm/cpu_mf.h>
 
 struct hws_ssctl_request_block     /* SET SAMPLING CONTROLS req block   */
 { /* bytes 0 - 7  Bit(s) */
@@ -68,36 +48,6 @@ struct hws_cpu_buffer {
 	unsigned int stop_mode:1;
 };
 
-struct hws_data_entry {
-	unsigned int def:16;        /* 0-15  Data Entry Format           */
-	unsigned int R:4;           /* 16-19 reserved                    */
-	unsigned int U:4;           /* 20-23 Number of unique instruct.  */
-	unsigned int z:2;           /* zeros                             */
-	unsigned int T:1;           /* 26 PSW DAT mode                   */
-	unsigned int W:1;           /* 27 PSW wait state                 */
-	unsigned int P:1;           /* 28 PSW Problem state              */
-	unsigned int AS:2;          /* 29-30 PSW address-space control   */
-	unsigned int I:1;           /* 31 entry valid or invalid         */
-	unsigned int:16;
-	unsigned int prim_asn:16;   /* primary ASN                       */
-	unsigned long long ia;      /* Instruction Address               */
-	unsigned long long lpp;     /* Logical-Partition Program Param.  */
-	unsigned long long vpp;     /* Virtual-Machine Program Param.    */
-};
-
-struct hws_trailer_entry {
-	unsigned int f:1;           /* 0 - Block Full Indicator          */
-	unsigned int a:1;           /* 1 - Alert request control         */
-	unsigned long:62;           /* 2 - 63: Reserved                  */
-	unsigned long overflow;     /* 64 - sample Overflow count        */
-	unsigned long timestamp;    /* 16 - time-stamp                   */
-	unsigned long timestamp1;   /*                                   */
-	unsigned long reserved1;    /* 32 -Reserved                      */
-	unsigned long reserved2;    /*                                   */
-	unsigned long progusage1;   /* 48 - reserved for programming use */
-	unsigned long progusage2;   /*                                   */
-};
-
 int hwsampler_setup(void);
 int hwsampler_shutdown(void);
 int hwsampler_allocate(unsigned long sdbt, unsigned long sdb);
diff --git a/arch/s390/oprofile/init.c b/arch/s390/oprofile/init.c
index 6efc18b5e60..9ffe645d598 100644
--- a/arch/s390/oprofile/init.c
+++ b/arch/s390/oprofile/init.c
@@ -1,19 +1,21 @@
-/**
- * arch/s390/oprofile/init.c
- *
+/*
  * S390 Version
- *   Copyright (C) 2003 IBM Deutschland Entwicklung GmbH, IBM Corporation
+ *   Copyright IBM Corp. 2002, 2011
  *   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
  */
 
 #include <linux/oprofile.h>
+#include <linux/perf_event.h>
 #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 +24,7 @@ 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
 
@@ -35,27 +38,67 @@ 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;
+
+	retval = hwsampler_allocate(oprofile_sdbt_blocks, oprofile_sdb_blocks);
+	if (retval)
+		return retval;
+
+	retval = hwsampler_start_all(oprofile_hw_interval);
+	if (retval)
+		hwsampler_deallocate();
+
+	return retval;
+}
+
 static int oprofile_hwsampler_start(void)
 {
 	int retval;
 
-	hwsampler_running = hwsampler_file;
+	hwsampler_running = hwsampler_enabled;
 
 	if (!hwsampler_running)
 		return timer_ops.start();
 
-	retval = hwsampler_allocate(oprofile_sdbt_blocks, oprofile_sdb_blocks);
+	retval = perf_reserve_sampling();
 	if (retval)
 		return retval;
 
-	retval = hwsampler_start_all(oprofile_hw_interval);
+	retval = __oprofile_hwsampler_start();
 	if (retval)
-		hwsampler_deallocate();
+		perf_release_sampling();
 
 	return retval;
 }
@@ -69,13 +112,20 @@ static void oprofile_hwsampler_stop(void)
 
 	hwsampler_stop_all();
 	hwsampler_deallocate();
+	perf_release_sampling();
 	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,
@@ -88,9 +138,12 @@ static ssize_t hwsampler_write(struct file *file, char const __user *buf,
 		return -EINVAL;
 
 	retval = oprofilefs_ulong_from_user(&val, buf, count);
-	if (retval)
+	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 +152,7 @@ static ssize_t hwsampler_write(struct file *file, char const __user *buf,
 		 */
 		return -EBUSY;
 
-	hwsampler_file = val;
+	hwsampler_enabled = val;
 
 	return count;
 }
@@ -109,38 +162,311 @@ static const struct file_operations hwsampler_fops = {
 	.write		= hwsampler_write,
 };
 
-static int oprofile_create_hwsampling_files(struct super_block *sb,
-						struct dentry *root)
+/*
+ * 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)
 {
-	struct dentry *hw_dir;
+	return oprofilefs_ulong_to_user(oprofile_hw_interval, buf,
+					count, offset);
+}
 
-	/* reinitialize default values */
-	hwsampler_file = 1;
+static ssize_t hw_interval_write(struct file *file, char const __user *buf,
+				 size_t count, loff_t *offset)
+{
+	unsigned long val;
+	int retval;
 
-	hw_dir = oprofilefs_mkdir(sb, root, "hwsampling");
-	if (!hw_dir)
+	if (*offset)
 		return -EINVAL;
+	retval = oprofilefs_ulong_from_user(&val, buf, count);
+	if (retval <= 0)
+		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;
 
-	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);
+	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 <= 0)
+		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 <= 0)
+		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 <= 0)
+		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 <= 0)
+		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 dentry *root)
+{
+	struct dentry *dir;
+
+	dir = oprofilefs_mkdir(root, "timer");
+	if (!dir)
+		return -EINVAL;
+
+	oprofilefs_create_file(dir, "enabled", &timer_enabled_fops);
+
+	if (!hwsampler_available)
+		return 0;
+
+	/* reinitialize default values */
+	hwsampler_enabled = 1;
+	counter_config.kernel = 1;
+	counter_config.user = 1;
+
+	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(root, "0");
+		if (!dir)
+			return -EINVAL;
+
+		oprofilefs_create_file(dir, "enabled", &hwsampler_fops);
+		oprofilefs_create_file(dir, "event", &zero_fops);
+		oprofilefs_create_file(dir, "count", &hw_interval_fops);
+		oprofilefs_create_file(dir, "unit_mask", &zero_fops);
+		oprofilefs_create_file(dir, "kernel", &kernel_fops);
+		oprofilefs_create_file(dir, "user", &user_fops);
+		oprofilefs_create_ulong(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(root, "hwsampling");
+		if (!dir)
+			return -EINVAL;
+
+		oprofilefs_create_file(dir, "hwsampler",
+				       &hwsampler_fops);
+		oprofilefs_create_file(dir, "hw_interval",
+				       &hw_interval_fops);
+		oprofilefs_create_ro_ulong(dir, "hw_min_interval",
+					   &oprofile_min_interval);
+		oprofilefs_create_ro_ulong(dir, "hw_max_interval",
+					   &oprofile_max_interval);
+		oprofilefs_create_ulong(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;
+		case 0x2827: case 0x2828: ops->cpu_type = "s390/zEC12"; 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 +481,17 @@ 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 +502,15 @@ 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 00000000000..61b2531eef1
--- /dev/null
+++ b/arch/s390/oprofile/op_counter.h
@@ -0,0 +1,21 @@
+/*
+ *   Copyright IBM Corp. 2011
+ *   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 */